SCROLL NEWS / DISCOVERY
Search headlines
Find the stories shaping the conversation.
Results for " measuring " 95 found 🔀 AI-powered Shuffle
SIA in-flight entertainment to get major upgrade, including larger screens, Bluetooth pairing
SINGAPORE – Travellers can expect a personal movie theatre experience on Singapore Airlines (SIA) routes after upgrades to its in-flight entertainment system, which will include touchscreens measuring...
More output isn’t better business: Workday’s CFO Zane Rowe on how finance chiefs should really measure AI
Zane Rowe discusses the company’s approach to AI agents, finance transformation, and measuring the impact of new technologies.
Could Prediction Markets Be a New Hedge Against Business Risk?
Could prediction markets become a new tool for managing business risk? Event-based contracts may help firms hedge revenue uncertainty, but success depends on accurately measuring exposure, matching risks and avoiding conflicts of interest.
Astronomers uncover new secrets of the Universe’s first five minutes
Scientists have achieved a major breakthrough in understanding the Universe’s earliest moments, measuring the amount of helium formed in the first five minutes after the Big Bang with unprecedented precision. Using one of the world’s most powerful telescopes, an international team studied 15 remarkably pristine galaxies that act as cosmic time capsules, preserving clues from nearly 13.8 billion years ago.
Pedri: Lamine Yamal shouldn't be compared to Lionel Messi
Pedri believes Lamine Yamal should not be compared to anyone, insisting that measuring the youngster against Lionel Messi is "complicated" because it means comparing him to "the best player in history...
Trump's new pitch for Republican candidates? They're really good looking
Aesthetics have long been part of the president's measuring stick, either as a source of praise or target for ridicule.
New Withings Users Must Pay $99/Year for Apple Health Syncing
Withings, the maker of smart scales for measuring body weight, has announced that customers who create an account on October 13, 2026 or later must have a Withings+ subscription for automatic syncing of data to third-party apps, including Apple's Health app. In the U.S., Withings+ costs $99.50 per year. If you created your account on or before October 12, 2026, automatic syncing to third-party apps remains free.
In Good Health: Are We Solving Hearing Loss?
For those experiencing hearing loss, being able to discern the speech of other people in noisy environments can be frustrating. And hearing aids often don’t convey sound in every situation.Hearing loss affects 37.5 million adults in this country. And rates of hearing loss increase with age. Some 10 percent of those over the age of 55 have hearing loss that qualifies as a disability. And just over half of all people over the age of 75 have disabling hearing loss.In this edition of “In Good Health,” we talk about the health issues that most affect us and put your questions to the experts. This time, we tackle hearing loss.Here are some resources shared by our experts during this discussion:- HearAdvisor.com – a hearing aid comparison and review site- Hear & Beyond by Shari Eberts and Gael Hannan – a guide to navigating hearing loss- Hearing Number app – a free smartphone app by Johns Hopkins University for measuring how much sound you can hearFind more of our programs online. Listen to 1A sponsor-free by signing up for 1A+ at plus.npr.org/the1a.
New sensor developed at Oregon State shows promise for fast, low-cost food-quality testing
CORVALLIS, Ore. – Researchers in the Oregon State University College of Science have developed a new electrochemical sensor for rapidly and accurately detecting a naturally occurring compound widely used in the food, beverage and fragrance industries.The compound is eugenol, which gives cloves their telltale flavor and smell, and measuring its presence in commercial products is important for manufacturers and regulators.
USWNT Gets a Measuring Stick Ahead of the World Cup
The USWNT is still finding its footing after falling to No. 1 Spain.
Washington’s Physicality Gap Amplified in Iowa Loss
Elijah Brown’s second-half performance struck life into the Washington offense, but it wasn’t enough to overcome the hole it dug itself against Iowa in the first half. The Huskies began the game with five straight punts. In what Jedd Fisch called a measuring stick game against the Hawkeyes, his team failed to measure up to …
The 2026 Nobel Prize in Physiology Went to a Technique That Lets Scientists Control the Brain With Light
Neuroscientific advances follow a pattern of visualizing the brain, measuring its function and then controlling it. Optogenetics may lead to a new era of medical treatment.
Measuring Stick for the Washington Run Defense Against Iowa
Two years ago, Iowa rushed for 220 yards on the Washington run defense, averaging nearly six yards per carry inside Kinnick Stadium. The Huskies lost 40-16 in what was just the fourth Big Ten game of Jedd Fisch’s tenure in Seattle. The loss demonstrated the physicality and stylistic gap that Washington aimed to close as …
Peng Gao receives NIH New Innovator Award for research measuring chemical exposure in blood and tissues | Harvard T.H. Chan School of Public Health
Peng Gao was awarded an NIH Director's New Innovator Award from the NIH's High-Risk, High-Reward Research program.
The challenge of measuring clinical value in mental health
Clients need assurances that high-cost medications flooding the system are aligned with the clinical support needed to improve mental health.
Experience Intelligence: 3 Insights on Measuring Business Impact
A new methodology is helping brands like IBM design for what matters most. When an industry standard doesn’t exist, create one. It’s the mentality of a
90 Ideas in 90 Minutes: Blaire Massa
Measure belonging, not just outputCompanies are exceptionally good at measuring transactions. At Ballet Des Moines, we know exactly how many tickets we sell, how many students attend our programs, how many dollars we raise. But those numbers look backward, not forward. We wanted to find out whether people were going to come back. So we…
Quantum teleportation breakthrough: Scientists crack a 25-year entanglement challenge
Scientists have developed and experimentally demonstrated a long-sought method for identifying W states, an important form of multi-photon quantum entanglement. The technique could make complex entangled systems much easier to measure, opening new possibilities for quantum teleportation, communication, and computing.
New soil test could reveal how soil microbes help feed plants
Scientists have streamlined a method for measuring DNA-bound phosphorus in soil, making it cheaper and easier while preserving its accuracy. The technique could reveal how soil microbes help recycle phosphorus and ultimately support more sustainable fertilizer and nutrient management.
Measuring the toll of extreme heat on workers’ health | Harvard T.H. Chan School of Public Health
Researchers are working to quantify the impact of working in extreme heat and humidity in order to help identify interventions that make a difference.
Eyes on Seagrass: Inside the Citizen Science Program
The Eyes on Seagrass Citizen Science Program was made for volunteers to help survey the bays, gathering crucial data to go towards measuring water health.
Monday Night Therapy: Huskers vs. Spartans Week 4
Minnie Hunt welcomes Husker Chuck back to Monday Night Therapy to recap Nebraska’s 3-0 start and look ahead to the Huskers’ first Big Ten matchup against Michigan State. While Nebraska handled its first three opponents as expected, Chuck cautions that conference play represents a significant jump in competition and should provide a much better measuring […]
The Big Holes in Wearable Heart Rate Variability And Readiness Scores
On September 9th, Apple announced it was revamping its Apple Watch Health Sensing System, rolling out a Readiness score (0 to 10), and increasing the frequency of heart rate variability (HRV) outputs 24-fold. This can be viewed as upping its competition with various other consumer wearable sensors. These “readiness scores” as a composite of multiple metrics, with heart rate variability (HRV) being front and center for most. The majority of Americans are now using wearable sensors, which equates to well over 100 million adults. HRV and Readiness scores are increasingly being marketed as a measurement of autonomic nervous system health, a digital marker for future disease, a clock for biological age, and a holistic metric to promote healthspan and even longevity. (Apple also introduced a new longevity tab and “Health Age”.) None of this has been proven. In this edition of Ground Truths I am going to review what we know about heart rate variability and readiness scores. Heart Rate Variability HRV is the variation in normal heart cycle timing. The variability of the heart rate, the barely perceptible millisecond changes in time between consecutive heart beats (see R-R intervals in the Figure below, left panel), is due to interplay between the sympathetic and parasympathetic (vagal nerve) inputs. Distinct from heart rate, individuals with the same heart rate can have very different HRVs. It is a rough reflection of the autonomic nervous system (ANS) activity, inadequate to say whether a person’s ANS function is abnormal. For more than three decades, heart rate variability (HRV) has been measured and several studies have found an association of low HRV and clinical outcomes, particularly a link with higher all-cause and cardiovascular mortality. There have also been less well established links of low HRV to risk of early cognitive impairment, dementia, mental illness, Type 2 diabetes and substance abuse. An important reminder is that HRV is a surrogate marker without any established cause-and-effect relationship. If you increase your HRV, that doesn’t mean it will improve health outcomes. In fact, there is no hard evidence for that. All that work linking to health outcomes was done with electrocardiogram (ECG) derived HRV. Now, in the era of consumer wearables, this is getting assessed differently, by optical pulse (yes, the lights you see) plethysmography (PPG) or what is called pulse rate variability (PRV). They are not the same, as shown below (right panel) and only concordant when the delay between the ECG and pulse is kept constant, which basically means at rest. I should mention there’s also what I will call MPV, a mattress mechanical movement sensor, a derived heart rate variability, that companies like Eight Sleep use, even further away from directly measuring HRV. HRV has not one uniform measurement but many different types of quantification, such as RMSSD, the magnitude of difference between successive R-R intervals of normal sinus beats (N-N) or SDNN, the standard deviation of NN intervals, both in milliseconds. SDNN is one of the so-called frequency domain HRVs (others are LF, HF, LF/HF). Different wearable sensors use different metics; Apple has relied on SDNN and nearly all of the others use RMSSD, which is generally considered the more accurate metric. There’s also the different length of time measured, such as for a matter of minutes, all day, or an overnight’s sleep. Short measurements are especially problematic since they don’t capture enough of respiratory modulation and other factors that influence HRV. Share Ground Truths How well does HRV correlate with PRV? There are very limited studies, especially independently done. One that is commonly cited was conducted by Air Force researchers in only 13 healthy adults assessing Oura ring 3 and 4, Whoop 4.0, and Garmin Fenix 6 and showed a correlation coefficient of 0.88 to 0.97 and a mean absolute percentage error from 6 to 10%. The correlation is not a perfect 1.0, but there’s at least a fairly high level of correlation. HRV is supposed to increase during the night due to takeover of the parasympathetic nervous system, and higher during deep sleep. A recent example of my 1 week, all day “HRV,” and one during sleep is shown below. As you can see, the N of 1 data are inconsistent for the same days from different sensors (Oura, AppleWatch, Fitbit Air, and Eight Sleep) by patterns, absolute numbers, and comparison with prior days and weeks. The largest study in over 8 million Fitbit users (the old version, not Google Fitbit Air, introduced in May 2026) gives you a sense of the effect of age, sex, and the 2 different main HRV (here PRV) metrics, with RMSSD on the left and SDRR (=SDNN) on the right below. That study, from data collected in 2018, is a major outlier, since all the more recent ones are tiny with respect to sample size. Many of the companies have not had independent evaluation of their HRV, such as Eight Sleep, but have published a low standard error on their website. There are some other published studies on the correlation between HRV and PRV, but they are all small and only in healthy adults. A scoping review emphasized the lack of study in underrepresented individuals, including the aged, people of color (which affects the PPG signal), and individuals who are underweight or obese. Add the typical adult age 60 plus with one or more chronic diseases. For example, one study in over 900 adults found poor correlation of HRV and PRV, non-uniformly underestimated across many chronic diseases (cardiovascular, endocrine, neurological, respiratory, and others), concluding PRV is “an invalid surrogate for HRV.” A recent systematic review of 43 studies comparing HRV and PRV found reasonable pooled absolute standardized error (HRV as gold standard) but only 10 of the studies provided quantitative synthesis in ideal conditions. Their main conclusion was similarly cautious: “PPG-derived HRV [PRV] should not be regarded as universally interchangeable with ECG-derived HRV across all devices, populations, and recording contexts.” Factors Affecting HRV and PRV That gets me to the long list of factors that affect HRV (and PRV) besides the device, the type of measurement (RMSSD, SDNN or others), the person’s signal, the sensor site, the duration of data capture, if weighting by sleep stage is used, how artifact is processed and corrected. And this list is not complete!: Oura puts out data from their community of users (who input data) on what affects their overnight HRV. The factors currently provided are: no alcohol (increase 8%), melatonin (increase 2%, float tank (increase 2%), wine (decrease 4%), and party (decrease 14%) in overnight HRV. Must be some big parties! Subscribe What is a PRV measurement good for? It has been falsely characterized as an index of “autonomic balance” and a specific indicator of stress. A 2018 review of the studies available for HRV and its relationship to stress, not using any of the current wearables, found that stress can lower HRV. But so can many other factors. The non-specificity of the signal, indexed to the table above, is striking. Evidence from a UK Biobank study of over 46,000 participants with actual HRV looked at genetically predicted HRV, a genetic risk score, that failed to show the expected HRV-mortality link, indicating that _HRV is likely not causa_l, but rather a reflection of person’s physiologic state. A review of consumer wearable HRV data from 5 longitudinal studies showed that nighttime PRV was not associated with perceived stress, and surprisingly higher HRV, in the largest cohort (N=717 participants), was correlated with higher stress. An Oura ring cohort of 525 first-year college students found a link between overnight PRV and perceived stress, but that was also seen with resting heart rate, sleep, and respiratory rate. Several very small studies have examined the relationship of HRV and athletic injuries or guiding training with mixed, and predominantly negative results. HRV biofeedback training with paced breathing had no significant effect on reducing stress or raising HRV, as demonstrated with sham controlled trials. When HRV for multiple days showed a decline in conjunction with body temperature, the Oura ring published data for prediction of Covid. The WHOOP company sponsored an observational study, published in 2026, of 30,000 users for 72 weeks, without a control group, that reported reduced alcohol intake (5.8 % points) by self-report. That doesn’t tell us much, and particularly about the merits of HRV for behavioral change. If you use the same device and conditions as longitudinal trends for multiple (at least 2-3) weeks that may be the one way to get something useful from the measurements. Data for overnight sleep with minimal motion and using RMSSD is the best proxy for real HRV. The reason to look at trends rather than any given night is that it more likely represents something, even though you won’t know with certainty what the “it” is. Keep in mind there are no data, no peer-reviewed evidence, to show that HRV fluctuation in-person has any correlation with health outcomes. Share Readiness Scores These are proprietary scores that integrate different metrics for each of the wearables: no algorithms have been disclosed. They are unvalidated against health outcomes. In a review of 14 composite health scores of readiness and recovery, HRV contributed 86% to the scores, followed by reading heart rate (79%), physical activity and sleep duration (both at 71%). That review noted the substantial variability in measurement protools and lack of standardization. Sleep staging is notoriously inconsistent and inaccurate by these sensors, which adds further to the HRV uncertainties for what the scores, which use sleep stage data, mean. Only resting heart rate has been shown consistently across devices to be extremely accurate. I’ve made a Table to summarize what we know about which metrics are included, the scores, any peer-reviewed studies that compared the readiness score with health outcomes, and the corresponding (if any, NA-not available) citation. You will note that some companies do not use the term “readiness," such as WHOOP for recovery, and Garmin, which has 2 different scores, one of which is Body Battery. Eight Sleep uses the term “Fitness Score.” They all include HRV; Apple includes a new metric they call “Recovery HRV” which among other components uses 7-days of sleep, but it is unclear what this means or how it is differentiated from other scores (there are clearly no data for outcomes). We have no knowledge of how the different components are weighted or whether any of these scores are better than resting heart rate, HRV alone, physical activity, or any other single metric. Since none of these are standardized, they are not interchangeable, so if you get a 90 for Oura that has no relationship to a 90 on a Google Fitbit Air. Notably, the company can update its algorithm for readiness score at any point without notification to device users. Without any useful evidence of actionability for these scores or established relationship with health outcomes, it is hard to make a case for their value. At the Apple recent announcement they showed their Readiness score (0-10) on the watch (Figure below) but there are no published data on this score, not even on their website. It’s available only on their new Watch Series 12 or Ultra 4 [of course, ;-)]. That exemplifies the problems with these scores, lack of data and evidence for being meaningful to promote health. Perhaps the best study (which isn’t saying much) is the WHOOP Recovery for golfer performance, because it did correlate with an objective outcome, even though there was no control group and the authors were all from the company. Among the 389 pro golfers, an absolute 10-per cent point increase in Recovery score was associated with about 0.5 fewer strokes per round. But that’s hardly a health outcome! WHOOP is also conducting a study in over 2,700 runners to see if their recovery score will be linked to less injuries and improved performance, but that is not yet published and has no control group or randomization. Putting This in Context For two decades I’ve been enthusiastic about the potential for digital health and particularly wearable biosensors. Over the years, we’ve seen some great progress for their ability to promote physical activity and accurately detect atrial fibrillation (the first FDA cleared deep learning AI for consumers). That work was the subject of rigorous research. But there are holes in the data and evidence for other metrics. One notable one is the “VO2 max” story that I wrote about earlier this year. At that time many subscribers asked me to cover heart rate variability, which I finally got to here. When I dived into the research and publication for HRV and readiness scores, I expected to find at least some that were of high quality and demonstrated their utility by linkage to health outcomes. To my surprise, I found none. The wearable sensor measurements for HRV (PRV) are, for the most part, accurate, but that validation work has only been done in small studies of healthy adults and does not take into account the long list of factors, from the device, software side, and the user side, that affect HRV measurements. Moreover, this metric chiefly relies on optical sensing and, as we have learned for heart rate PPG sensing, may be less accurate in people of color. Keep in mind that all of the health outcome association evidence comes from ECG-derived HRV; none are from wearable sensor data. I will repeat the key point: there's no peer-reviewed evidence to show that in-person HRV fluctuation—or efforts to raise your HRV— has any correlation with health outcomes. For those of you who look at your HRV on awakening, or even 2+ week trends of it being low, I hope this context helps to relieve any anxiety. Yes, low HRV (not PRV) has been shown to increase risk of some diseases as summarized above. But efforts to raise your HRV—a surrogate metric— has not been established for improving any health outcomes. HRV does not have any evidence of causality (the genetic evidence actually goes against this possibility). In this summary, I have not included data for other wearable sensors such as Polar, Samsung, Withings, Suunto, Amazfit, Coros, Ultrahuman, or additional mattress sensors. These are beyond my first-hand experience, and as far as I know from my in-depth review none have any peer-reviewed published data that differ from the 6 sensors I’ve reviewed here. From the points I’ve gone over above, we’re not ready for readiness scores. Besides being proprietary, they are predominantly based on metrics that have their own issues. It’s compounding the problem, like building a house without a solid foundation that has never undergone a rigorous inspection, and then selling it. Like I mentioned for PRV, you can look at trends over weeks rather than any single day, to get a handle, but even that may not be helpful. There’s simply no evidence that these scores meaningfully relate to health outcomes. I’d emphasize they might, but that requires doing prospective or randomized studies to prove it. None exist. There’s great promise for HRV/PRV utility_. For example**,**_ Prof Maiken Nedergaard, who discovered the brain glymphatics that are essential in eliminating metabolic waste products from the brain during sleep, has posited that HRV could be a non-invasive marker for neuromodulator oscillations, brain-body regulatory circuits, and brain clearance. That would be extremely useful, but like everything else on HRV and readiness scores it requires solid research and validation. The lay media isn’t helping much to get the story straight. Earlier this year The Economist published a piece entitled “The most useful indicator of your overall health” which ordained HRV as an “accumulated stress score.” That’s akin to the false assertion about VO2max: “V02 max is the singular most powerful marker for longevity.” As I’ve summarized here, that is not established. The fact is that so many things can lower HRV, including physical exercise (especially an intense workout), reduced sleep quality, stress, the list above, no less the device, signal, and software. Whatever fluctuations observed have not been correlated with any health outcome. Sadly, “datamaxxers” are widely using HRV and readiness scores that have never been validated to mean anything. We already know that for some people using the sensors for sleep metrics, it can induce “orthosomnia,” an obsession to get high sleep quality, with associated high levels of anxiety. In an experiment done by a company to promote sleep quality for its employees, “For those employees who did use the trackers, many reported feeling perfectly rested until their tracker told them they had had a terrible night. Others were told that they had slept like a baby when they had actually been lying awake worrying about the quality of their sleep. “ The same problem can result from preoccupation with HRV or readiness scores, with anxiety that would lead to further reduction in both. It you are using a wearable like >100 million American adults, it’s OK to look at these data, but contextualized with the major caveats reviewed here. If you are one to require evidence that HRV or readiness scores are linked to health outcomes, you may not even want to look. The companies make it hard to turn them off! Let me end with the companies that make and sell wearables. Apple’s doubling down on HRV (24-fold more reporting and heart rate very 5 seconds) and introduction of a Readiness score tells us that consumers have bought into these metrics and they are joining the club. However, all of this is occurring with a backdrop of tens millions of users, claims about the data that are not backed up by adequate evidence, marketing way out in front of whatever limited data exists, and not being transparent about their readiness score algorithms. The companies can well afford to do the research that is needed to connect these metrics with health outcomes show, once and for all, that increasing HRV or using readiness scores promotes our health. If they believed and invested in the products they are selling, we’d not be in this position of not knowing. That’s essentially where we are with HRV and readiness scores. Perhaps someday this will change and we’ll have good reason to embrace them. NB: I wrote this post. No AI. I have no conflicts of interest with any of its content. Loading... Ground Truths has 215,000 subscribers from every US state and 214 countries. There are over 300,000 followers of Ground Truths so more than 90,000 folks who can easily convert to be free subscribers. Your subscription to these free essays and podcasts makes my work in putting them together worthwhile. If you’re not a subscriber, please join! If you found this interesting PLEASE share it! Share Ground Truths The proceeds from all voluntary paid subscriptions go to support our summer internship program. It enabled us to accept and support a record number of 62 summer interns that joined us in 2026! These are high school, college and medical students selected from thousands of applicants. We couldn’t do this expanded program without the funds coming in through Ground Truths. Thank you!
Razorback Report Card: Georgia
The Razorbacks do need to show they can learn from an embarrassing performance. Georgia is not a reasonable measuring stick for a one-week turnaround, but Arkansas must show more than a respectable final score Saturday. The Razorbacks need answers in five specific areas after the 43-10 loss at Utah. Razorback Report Card: Georgia Five Questions …
Setting Expectations for Taylor Jenkins’ First Season as Bucks Coach
The new coach brings a winning resume to a rebuilt roster, so the fair measuring stick this season is growth rather than a playoff push.
Do black holes have hidden hair? Scientists are searching
If black holes have extra matter around them – aka hidden hair – astronomers could find it by measuring the different characteristics of these ripples.
Microbes—dead or alive—can be early indicators of soil health under regenerative farming
Researchers have demonstrated a series of biological tests that could serve as "proxies" for measuring how different regenerative agriculture practices could improve soil carbon—years before it would ...
Huawei's Watch D3 Doubles as an Inflatable Blood Pressure Cuff
The Huawei Watch D3 can track your sleep and workouts, but its best trick is a clever method for measuring your blood pressure.
Measuring ROI For Healthcare AI May Require A New Approach
AI tools in healthcare may add value in novel and non-traditional ways.
Lenovo Recognized as a World's Top Disability Inclusive Business Reflecting the Global Impact of its Disability Advantage Strategy
Lenovo today announced it has been recognized as a World's Top Disability Inclusive Business through the Disability Index®, the world's leading benchmarking tool for measuring and advancing disability inclusion in business
It’s time to measure what matters in higher education
Commentary: Dr. Ian Mortimer of Golisano Institute highlights the importance of measuring graduate employment and financial independence immediately after college in Rochester.
Sea Aggie Researchers Make Evening Fun with Science on the Beach
“The environment is where we all meet; where we all have a mutual interest; it’s the one thing all of us share.” – Lady Bird JohnsonOn Galveston Island, beaches rank among the favorite environments to share and study, and, as Texas A&M University at Galveston faculty emphasized last week, are “More Than Just Sand.”A program subtitled “The Secret Life & Value of Sandy Beaches” drew more than 40 people to the Aug. 13 meeting of Surfrider Foundation Galveston, a local chapter of a nonprofit focused on the protection and enjoyment of oceans and beaches.The evening began barefoot with ecology lessons on the sand before moving across Seawall Boulevard for beverages, bites, and more beach talk at an area establishment. Drs. Guilherme Corte and Yasmina Shah Esmaeili show how a mesh bag is used as a sieve to separate organisms from samples of the sand. Credit: TAMUG Divisions of Marketing and CommunicationsNot far from where others were enjoying the attractions at Pleasure Pier, Drs. Guilherme Corte and Yasmina Shah Esmaeili introduced the ecologically curious to species living within the sediments.“When people hear ‘coral reefs,’ they immediately think of fish, live ecosystems, but when they think of sandy beaches, they think party, recreation, vacations – not coastal ecosystems,” said Corte, an assistant professor of marine biology in the College of Marine Sciences and Maritime Studies at TAMUG.Shah Esmaeili co-founded the Gulf eDNA Network and works with Corte in TAMUG’s Benthic Ecology Lab.“Sandy beaches are indeed great places for leisure and recreation,” Corte said. “Let’s keep enjoying them, because they are awesome. But they are also full of life that we most of the time don’t realize is there.”The married couple soon had the beachcombers measuring the slope with PVC poles, searching for ghost shrimp with a stainless-steel pump, sifting for marine worms and sand fleas, and guessing which species would burrow fastest back into the sand once released.Away from the sand, the focus turned to research from Drs. Ashley Ross ’03, ’10, and Virgie Greb, ’25 and those words from Lady Bird Johnson reminding people that the environment isn’t separate from their daily lives.“It’s about things that we value, mostly our communities, strong local economies, outdoor recreation, cultural connections, and our overall quality of life,” said Greb, a proud BOI – born on the island – Galvestonian, who now works in Austin as a project manager for the Texas General Land Office. “Along our coast, these benefits are deeply interconnected.”Greb and Ross, an associate professor in marine and coastal environmental science with the College of Marine Sciences & Maritime Studies, have studied a variety of ways beaches and the sea are important to Galveston and those who live there. Drs. Ashley Ross (left) and Virgie Greb (standing at right) field questions after their presentation. Credit: TAMUG Division of Marketing and CommunicationsIn 2023, using three survey methods – online for Galveston County residents, in-person at beaches, and paper distributed via city water bills – they collected 732 complete responses, 53 percent from residents and 47 percent from visitors.Most were frequent beachgoers who gave the recreation value of Galveston beaches a “B” and rated amenities such as restaurants, lodging, shopping, and beach access highly. Lower rated were water clarity and restroom facilities.“It’s noticeable that primary residents actually rate water quality higher; they’re more tolerant of hard, turbid water,” Ross said, crediting them with better understanding the natural conditions with the Trinity and San Jacinto Rivers bringing sediments and rich nutrients to the area around Galveston.Island residents also were more likely to understand that “clear” isn’t the same as clean beach water and to know that water quality is regularly tested. Clean means absent of disease-causing pathogens.“We need more education and outreach focused on water quality and the benefits of Galveston beaches and shores and water,” Greb said. “And it’s not just for tourism. There are a lot of people who are very lucky to live on an island with beaches and the bay.”
No Neutral Number: The Politics of Defining and Measuring Inequality
Before anyone can reduce inequality, someone has to define it, decide how to measure it, and choose how to respond.
Researchers Find That Waist Size Is a Powerful Indicator of Health Risks
Measuring waist size may be one of the simplest and most effective ways to detect obesity-related health risks, according to a study by Rutgers Health researchers. The study, published in JAMA Network Open, found that combining body mass index (BMI) with waist circumference, or using waist circumference alone, performed as well as a newly proposed obesity definition that incorporates multiple body measurements.
Scientists solve the mystery of a brain “switch” that can trigger weight loss in opposite ways
Scientists have uncovered why two seemingly opposite ways of targeting the same brain receptor can both promote weight loss. In mice, activating the GIP receptor in the brainstem reduced appetite, while blocking it in the hypothalamus removed a kind of “brake” on fullness signals. The discovery helps explain why very different obesity drugs can achieve similar results and suggests that carefully combining GIP-targeting treatments with GLP-1 drugs like those related to Wegovy and Ozempic could produce stronger effects.
Measuring Whether Behavioral Health Crisis Systems Reach People in Need
Since passage of the American Rescue Plan Act in 2021, the United States has made significant investments in behavioral health crisis response systems,
3 main takeaways from Cowboys’ joint practice with Rams
The Dallas Cowboys got their first opportunity to test themselves against team during Tuesday’s joint practice with the Los Angeles Rams, and there was plenty to like — especially on defense. Facing reigning MVP Matthew Stafford and a Rams team off a trip to the NFC Championship Game provided Dallas with a strong measuring stick before preseason action. The Cowboys defense answered the challenge, creating turnovers and consistently disrupting Los Angeles’ offense.
Small Business Employment Index Picks Back Up
NFIB’s July Jobs Report shows that the Small Business Employment Index rose after four consecutive months of decline, registering 102.1 in July after measuring 100.2 in June. The current reading is above the 2025 average of 101.2 and the historical average of 100.0. In July, 36% (seasonally adjusted) of small business owners reported job openings they could
Bigger Displays Rumored for Next Year's iPhone Models
Apple is prototyping a second new iPhone display, this one measuring around 6.4 inches, that could be destined for the 20th anniversary iPhone, according to a Chinese leaker. In a recent post on Weibo, the leaker known as "Digital Chat Station" said Apple is currently sampling two new display panels.
Decoding the gut microbiome's symphony in health and disease
The trillions of microorganisms living in our intestine influence our health, from digestion and metabolism to immunity. Yet scientists still struggle to understand how these microbes shape these important physiological functions because gut microbiomes are extraordinarily complex, redundant, and dynamic. New technologies are therefore essential for advancing microbiome research. Researchers at the University of Vienna have now reported effective strategies for simultaneously measuring thousands
This once-a-week workout may help cut belly fat, study shows
A surprisingly small dose of exercise may deliver major health benefits for adults carrying excess fat around the waist. In a four-month trial involving 315 adults with central obesity, researchers found that completing 75 minutes of brisk interval walking in a single weekly session reduced body fat, trimmed waistlines, and improved cardiovascular fitness about as effectively as splitting the same exercise across three days.
Science Inquiry Comes Alive for Jacket Middle Schoolers
Snails, sidewalks and Double Stuff Oreos... At Partnership Middle School, instructors are finding innovative ways to help students connect the classroom to what they see in the real world every day. By integrating hands-on experiences and scientific investigations into weekly lessons, 6th and 7th grade science teachers push their classes to explore the curriculum standards through the lens of critical thinking, communication with their peers, experimentation, and scientific inquiry. Last week, two different approaches for two different grade levels saw Jacket teachers ignite the same spark for learning. On Tuesday, 6th graders in Lydia Jones' science class took their lessons outside as they explored the school grounds to identify living and non-living things. Despite the warm temperatures, they engaged in discovering unexpected objects on the campus and connecting that to what they learned in class about the criteria for what qualifies a thing as "living". Ms. Jones tasked students with walking the grounds and listing as many living and non-living things they could find in just a few minutes. Students spotted a fire hydrant, construction vehicles, grass, trees, soccer equipment, rocks, sidewalks, and even a snail hidden in the plants. After they finished the exploration, students returned to the regular classroom to discuss their results and whether their objects were classified correctly. On Thursday, 7th graders in Omini Parks' science class participated in their first lab of the semester as they explored and identified the physical properties of a favorite snack -- Oreos. Ms. Parks identified the class objective before the lab activity began by writing the academic standard on the dry erase board in a declarative statement: "I can collect and evaluate qualitative data to describe substances using physical properties." Students watched a video about how Oreo sandwich cookies are made, and then engaged in a question and answer discussion to give observations on the manufacturing process, review various types of physical properties and their scientific terminology, and talk generally about how the lab inquiry would work. The Q&A ended with a predictable question from students... "Will we be able to eat the Oreos?" The hands-on lab exercise required the 7th grade scientists to observe and record the differences between regular Oreos and the Double Stuff variety based on their physical properties. They followed the instructions in their lab brief to chart a comparison between the two types of cookies by observing properties with their magnifying lenses, measuring the height and diameter of both, and weighing the cookies with and without their filling. The activity let students put their lessons about physical properties into practice with real objects while also practicing vocabulary and writing skills to record their observations and using math skills from the last school year like taking measurements and calculating averages. Of course, they ended the lab by charting the taste of the cookies as well! These types of activities allow students to see the science standards they are learning come to life in the real world and engage with the content in new ways. Beyond that, they serve as a foundation for encouraging and growing scientific inquiry. Later in the school year, students will use the scientific method they are learning to complete science projects that require real research and the ability to explain and answer questions about their process and results. At the school-level competitive Science Fair last year, 41 Partnership students earned a place in the 2026 regional Science Fair held at MSU. From there, 14 advanced to the State Science Fair conducted at Northeast Mississippi Community College in April.
NASA’s Psyche spacecraft aces Mars flyby capturing stunning timelapse video
NASA’s Psyche spacecraft aced its Mars flyby, using the planet’s gravity to speed toward its 2029 encounter with a metal-rich asteroid. The spacecraft tested its cameras, magnetometer, and particle-detecting instruments, capturing unusual views of Mars and measuring its magnetic environment. It also detected neutrons from the planet and spotted the tiny moons Phobos and Deimos.
Breaking Down the Hard Business of Measuring Security Culture
By understanding the types of data available and how they can be meaningful, security professionals can develop metrics that are useful in describing and enhancing security culture.
Apple Testing 7-Inch Display for Largest 20th Anniversary iPhone Model
Apple is reportedly prototyping a larger iPhone display measuring roughly 6.96 inches, which would likely be marketed as a 7-inch panel if it reaches production. According to Chinese leaker Digital Chat Station, Apple is testing the display internally in preparation for the 20th anniversary iPhone, which is expected to launch next year.
Oxford Brain Diagnostics Advances Vision for Measuring Brain Health and Neurodegenerative Disease with Five New Studies at AAIC 2026
New findings demonstrate the potential of OBD's MRI-based Cortical Disarray Measurement (CDM®) technology to support patient identification, disease characterization, treatment monitoring, and therapeutic development across Alzheimer's disease OXFORD, United Kingdom, July 13, 2026 (GLOBE NEWSWIRE) -- Oxford Brain Diagnostics (OBD), a leader in MRI-based measurement of brain health and neurodegenerative disease, today announced the presentation of five new studies at the Alzheimer's Association I
Your resting heart rate could reveal more about your health than you think, doctors say
Here is what your resting heart rate reveals about cardiovascular health and when to see your doctor. Pulse checks and wearable devices can help with measuring the rate.
Born to Soar: First Baby Born at Orlando Health Watson Clinic Lakeland Highlands Hospital Receives Polk State Foundation Scholarship
When baby Liam entered the world at 11:44 p.m. on Wednesday, July 8, weighing 7 pounds 1 ounce and measuring 20 inches, he became the first baby born at Orlando Health Watson Clinic Lakeland Highlands…
How HCA Healthcare uses scale to improve patient care
The health system’s expansive network is designed for testing new ideas, measuring results and implementing proven solutions across all sites.
From chasing titles to building businesses: Tennis legend Rafa Nadal on his next act
From chasing titles to building businesses, 22-time Grand Slam champion Rafa Nadal speaks to CNBC’s Tania Bryer about retiring from professional tennis, his new chapter as an entrepreneur, and measuring success without trophies.
Small Business Job Openings Rebound
Texas Border Business AUSTIN, Texas – NFIB’s June Jobs Report shows that the Small Business Employment Index remained essentially flat, registering 100.2 in June after measuring 100.3 in May. This is the fourth consecutive month the Index declined. The current reading is below the 2025 average of 101.2 but still slightly above the historical average of 100.0. In June, […]
UC Davis to Set Standard for Measuring Airborne Nanoplastic Health Risks
UC Davis researchers have received a nearly $4 million grant NIEH to develop the world’s first standardized method for measuring and describing the neurotoxicity of airborne nanoplastics.
Meaningful Outcomes Determine the Winners of the Health AI Race
Global investment in healthcare AI exceeded $18 billion in 2025, yet health systems still lack a shared understanding of the outcomes that investment should deliver. This paper argues that success in the health AI race will be determined not by technological sophistication but by which systems first identify and pursue the outcomes that matter most to patients, professionals and health systems. Without a shared language for defining and measuring what healthcare AI should deliver, systems risk optimizing for what is easiest to count rather than what matters most. This paper provides a framework for ensuring AI investment translates into real value for healthcare systems and the populations they serve.
Efzimfotase alfa demonstrated improvements in bone health in treatment-naïve pediatric patients with hypophosphatasia achieving median difference of 1.67 vs placebo in RGI-C Score at week 25 in MULBERRY Phase III trial
Positive results from the MULBERRY Phase III trial showed that efzimfotase alfa (ALXN1850), an investigational enzyme replacement therapy, demonstrated a statistically significant and clinically meaningful improvement in bone health in children (2 to
From Linchpin to Lever — Part 1: Measuring Trust in Public Health
Learn how local health departments can measure public trust using existing public health tools to strengthen relationships and improve community health outcomes.
Giant viruses discovered on Greenland ice sheet may slow glacial melting - Futura-Sciences
Giant viruses defy microbiological norms Lead author Laura Perini and her team at the Department of Environmental Science at Aarhus University identified the microorganisms during a field survey of Arctic ice. Measuring up to 2.5 micrometers in length, these viral specimens are roughly 125 times larger than conventional viral strains....
Is SMIC N+3’s Metal Pitch Smaller than Intel 18A’s?
Almost four years ago, we published that SMIC had started shipping 7 nm (N+1) chips. Now, SMIC is shipping its third-generation 7 nm (N+3) process in Huawei’s Kirin 9030, with a minimum metal pitch of 32.5 nm, about 10% tighter than the 36 nm minimum metal pitch shipping in Intel’s latest Panther Lake CPUs on 18A. The headline is true, but incomplete cherry picked metric. N+3 reaches the density of TSMC N6 through aggressive DUV multi-patterning and design-technology co-optimization (DTCO), but it pays for that in complexity, efficiency and process control. We found this and more in our reverse engineering and teardown where we cover SMIC’s N3 process technology, Huawei’s packaging, memory, architecture, and more. SemiAnalysis has been building a state-of-the-art teardown lab in Oregon capable of analyzing the world’s most advanced and important chips over the last year and half. We have already generated revenue on advanced datacenter chip teardowns including our recent reverse engineering of a major TSMC customer’s COUPE CPO optical engine + EIC 3D stack. This is the first public report from the SemiAnalysis Teardown Engineering & Evaluation Lab, or STEEL for short. The lab is aggressively scaling up and out and we’re excited to announce it publicly. This is a bit of inconvenient timing for TechInsights as they are private equity owned and currently being sold while having enjoyed virtually no credible competition for decades. This has led to TechInsights underinvesting in CAPEX. SemiAnalysis exceeds TechInsights in revenue despite no venture or private equity ownership and being founded only 6 years ago. Because we have no external investors and are founder led, we move faster, build faster, and we can release client chip teardowns for free regularly, while focusing on datacenter for our major clients. Here’s the first public image from our lab, the HiSilicon Kirin 9030 Pro SoC: This report will detail our teardown of the Kirin 9030 and our findings on SMIC’s N+3 process, the most advanced in China. For comparison, we’ll show our teardown of the MediaTek Helio G99, made on TSMC N6. Through this comparison, we can look at the effect of export controls – SMIC N+3 and TSMC N6 are comparable nodes, but one is heavily export-controlled, the other free to use the West’s most advanced equipment. Here we see both China’s progress and constraints. SMIC N+3 reaches TSMC N6-class logic density, but it requires far more aggressive DUV multi-patterning, so it does not match N6 on process maturity or cost. The Kirin 9030 Pro performs similarly to three-year-old Android flagships, and trails far behind the current flagship SoCs from Apple, Qualcomm, MediaTek, and Samsung. The efficiency gap is even wider. Export controls have not stopped Huawei and SMIC from shipping advanced silicon, but they have forced a different path. Without EUV, SMIC is leaning harder on DUV multi-patterning, DTCO, and increasingly complex integration. The roadmap continues forward through tighter design rules and backside power, but each step adds cost and process risk. Huawei’s τ scaling and LogicFolding show another path: stacking active logic and recovering density through advanced packaging and system-technology co-optimization (STCO). To understand the Kirin 9030, we must first understand Huawei’s SoC history. HiSilicon is Huawei’s chip design arm, responsible for the Kirin smartphone SoCs, Kunpeng server CPUs, Ascend AI accelerators, and switch/router networking silicon. Before export controls, Huawei was TSMC’s largest customer – the only customer on TSMC’s first EUV node, N7+, and among the first on N5, alongside Apple. That ended in late 2020. Huawei switched to Qualcomm SoCs in its flagship smartphones, though export controls limited them to 4G-only variants. In late 2023, Huawei returned to in-house silicon with the Kirin 9000s, a successor to the Kirin 9000, fabricated on SMIC N+2 instead of TSMC N5. In the following years, they released the Kirin 9010 and 9020 on the same N+2 process. These chips used Huawei’s in-house TaiShan CPU cores and Maleoon GPU. We have not torn down a Kirin 9020 ourselves, so the predecessor die shot is from Kurnal. The die shots show how Huawei spent its silicon budget: which functional blocks are where, and how their areas compare to the predecessor. First, a quick guide to the major blocks on the die. The total die area is nearly identical, but the 9030 uses that area more aggressively. A denser process lets Huawei fit an extra middle CPU core, more GPU and NPU cores, and larger caches into the same footprint. In contrast, the Helio G99 is a much smaller, low-cost SoC, built for budget smartphones rather than a flagship device. While the Kirin 9030 is ~140 mm², the G99 is only ~29 mm², roughly one-fifth the area. The underlying TSMC process technology, however, is directly comparable as a baseline for analyzing SMIC’s. The Kirin 9030 is an evolutionary refresh, not a clean-sheet design. Its CPU, GPU and NPU cores carry over the 9020’s families, and the gains come from three levers: the SMIC N+2-to-N+3 process step, DTCO and floorplan work, and incremental microarchitecture. Area is where the first two show up, and the 9030 scales well here. Performance and efficiency are the harder test. Huawei’s design holds up better than its node would suggest, but the chip still trails, both because N+3 sits behind the leading-edge nodes and because its cores, while competent, remain a few generations behind the newest designs. The new prime core is an incremental update. The main changes are a 10% frequency increase from 2.5 GHz to 2.75 GHz and a doubling in the L2 cache from 1 MiB to 2 MiB. Despite the increased cache, the core size decreased by 7.6%. Excluding the private L2 cache, the core size decreased by 21%. This is a large reduction for an incremental node. Compared with the TaiShan New V120 core in the Kirin 9020, the Kirin 9030’s middle core is almost unchanged architecturally, yet each core shrinks by ~22%. Most of that comes from the move from N+2 to N+3, with layout likely accounting for the rest. Visually, the most noticeable change is the increase from 3 middle cores to 4. There is also a 20% increase in the shared L3 cache of the big cluster. This helps to improve multi-core performance without sacrificing much in terms of area. Even with each core shrinking, the big CPU cluster’s total area is essentially unchanged. The per-core savings went back into an additional middle core and larger caches. The tiny cores shrank less than the prime core (excluding its L2 cache) and less than the middle cores. This is likely because fixed overhead is a larger share on a small core. We cannot resolve any architectural changes from the die shot alone, but the per-clock and efficiency gains shown below point to more than pure process and layout scaling. The area reduction was offset by a doubling of the shared L2 cache from 2 MiB to 4 MiB, leaving the total tiny CPU cluster area slightly larger. Area is the easiest improvement to see from a die shot, but it is only one part of PPA (power, performance, area). For modern logic, power and performance matter just as much, and often more. Since Dennard scaling broke down in the mid-2000s, voltage and frequency have not scaled in step with transistor dimensions, so each node has had to fight harder for gains in performance and efficiency. The starkest comparison is not Kirin 9020 versus Kirin 9030 Pro. Apple’s efficiency cores run circles around Huawei’s prime core. Apple’s low-power core delivers 20% higher integer performance while drawing only 1 W, compared with 4.5 W for Huawei’s prime core. N+3 matches TSMC N6, but N6 is several generations old. Apple and Qualcomm build on N4 and N3P, which are denser and sit on a better voltage-frequency curve, giving them a larger transistor budget and more performance per watt. The 9030’s own cores did improve. The middle and tiny cores gained 17% and 14% in per-clock integer performance over the 9020, with floating-point flat on the middle core and up 11% on the tiny. The tiny core improves cleanly, with performance rising while power falls and efficiency increases by 45% in integer and 24% in floating point. The middle core is mixed: integer performance rises but power rises faster, cutting integer efficiency by 7%, while lower power lifts floating-point efficiency 16%. Per-clock gains at the same or lower frequency are microarchitectural, so the cores are tuned, not just shrunk. Both also failed to hold their rated maximum frequencies, pointing to thermal, power, or stability limits. Per clock, the middle core sits around Arm Cortex-A720 and the tiny core near the Cortex-A520; absolute performance trails because Huawei clocks them much lower. The prime core is roughly Cortex-X2 class per clock, a 2021 design. Apple’s 2020 M1 Firestorm core is still 35% higher per clock and 57% faster in absolute integer performance at a similar 4.5 W. The current leading edge is further ahead again: the Apple M5 P-core is 60% higher per clock and 2.7× faster, the Arm C1 Ultra 45% higher and 2× faster. Matching older high-end cores per clock is a genuine design achievement. What Huawei cannot match is the voltage-frequency curve and transistor budget of leading-edge nodes, which let Apple, Qualcomm and others spend more transistors in the same area on wider cores, larger caches and deeper buffers while running at lower voltage. Huawei’s LogicFolding roadmap is one answer, stacking active logic to recover density and shorten signal paths. We return to it later. The GPU compute units (CUs) changed more visibly than the CPU cores, moving to a more rectangular layout for both the arithmetic logic unit (ALU) clusters and the CU overall. Even with ray-tracing support added, a CU shrank ~28%. However, that shrink is offset by the increase from 4 to 6 CUs and the area outside the CUs grew 33%. Overall, the GPU cluster is larger by ~10%. The GPU is where Huawei makes its biggest gains. The Maleoon 935 is not competitive with current flagships, but it is a large step up from the 920 and reaches older-flagship territory. In 3DMark it is 70% faster in Wild Life Extreme (WLE) and 79% faster in Steel Nomad Light (SNL) than the 920; with 11% higher clocks and 50% more CUs, the ~67% theoretical uplift roughly matches WLE and is beaten by SNL. It edges ahead of the Snapdragon 8+ Gen 1 in WLE and SNL, and the Dimensity 9200 and Apple A16 in WLE, but stays far behind newer parts: the Snapdragon 8 Elite Gen 5 and Dimensity 9500 are ~2.4–2.6× faster in WLE and ~3.2× faster in SNL. The Maleoon 935 is Huawei’s first GPU with hardware-accelerated ray tracing; there it lands slightly ahead of the Exynos 2200, and on par with the Apple A16, with current flagships up to 3.7× faster. The Neural Processing Unit (NPU) saw the largest structural changes of any block, moving from a Lite and a Tiny core in the Kirin 9020 to a Lite and two Tiny cores in the Kirin 9030. Both core types also show significant layout changes. This is a reversal in Huawei’s NPU design. The Kirin 9000 5G, its last flagship chip on TSMC N5, used two Lite and one Tiny core. The series of SoCs on SMIC N+2 moved to one Lite and one Tiny core, likely for area savings. With the Kirin 9030, Huawei has shifted back toward a larger multi-core NPU cluster, but with the additional area going to a Tiny core rather than a Lite core. We’re diving deep into the most advanced datacenter and AI hardware hitting the market. To learn more about what’s in the pipeline or to commission a custom teardown, contact sales@semianalysis.com. Interested in joining us on this ride and think you can be a difference maker? Check out our Careers page. Before diving into the process stack, the package and memory are worth separating from the SoC itself. The Pro variant of the Kirin 9030 carries 12 GB of Samsung DRAM, with two stacks of four dies each. The dies were identified as the K4L2E165YD, a 12 Gb LPDDR5X-9600 device fabricated on Samsung’s 1a node, the fourth generation of its 10 nm-class DRAM after 1x, 1y and 1z. 1a has shipped in volume since 2022, so this is current memory rather than older-node inventory. The 16 GB Pro Max variants we obtained were found with both CXMT and Samsung packages. The CXMT package is marked CXDD7JEDM, with two stacks of four dies, packaged in week 45 of 2025. The inferred die dimensions from X-ray computed tomography (CT) are consistent with a known density of ~0.3 Gib/mm² for the CXMT G4 process, roughly equivalent to other manufacturers’ 1z processes. The Kirin 9030 uses a typical integrated package-on-package (iPoP) stack: multiple DRAM dies in a memory package sit above an organic redistribution layer (RDL) interposer, which sits above the SoC and package substrate. The full package is then mounted to the printed circuit board (PCB) through ball-grid array (BGA) solder bumps. The memory package substrate is a thin bismaleimide-triazine (BT) laminate carrying the LPDDR5X stack. The organic RDL interposer over the SoC routes the PoP signals around the die and carries possible dummy thermal copper pillars. The package substrate, a thicker Ajinomoto Build-up Film (ABF) build-up over a BT core, fans the flip-chip bumps out to BGA pitch and embeds the power planes. The whole stack is organic. The only silicon is the SoC and the LPDDR5X dies; there is no silicon interposer. Keeping it all-organic brings the package’s coefficient of thermal expansion (CTE) close to the PCB’s, reducing board-level warpage, and avoids the cost of a silicon interposer the SoC’s bandwidth does not need. In an iPoP stack, the memory package connects to the organic RDL interposer through an array of solder bumps. Underfill fills the gap around those bumps, adding stiffness and protecting the joints from mechanical stress. The Pro and Pro Max variants differ here, which we cover behind the paywall. The die shot and architecture tell us how Huawei allocated its silicon budget. The process tells us what SMIC can manufacture. We use the Helio G99 as the process reference for TSMC N6. Both SMIC N+3 and TSMC N6 are evolutions of previous 7 nm-class nodes. We used targeted TEM cross-sections through logic and memory regions, imaged in both fin-cut and gate-cut directions. Each cross-section caption gives its horizontal field width (HFW), the real width of the imaged area. We start at the transistor fins, then move up through standard cells, local interconnect, and SRAM. SMIC has not overtaken Intel or TSMC. It uses aggressive DUV scaling and DTCO to reach N6-class density, but that density doesn’t translate into comparable performance and efficiency, for two reasons: the node gap to leading-edge nodes, and Huawei’s core designs. Fin Profile One of the most important knobs in a FinFET process is the fin profile: the shape of an individual fin and the channel where current passes from source to drain. The ideal fin is tall, narrow, and nearly vertical. A taller fin increases effective channel width, while a narrower fin improves electrostatic control by thinning the body the gate must control. Push either too far, and the process pays for it: weaker drive current, fragile fins, taper, footing and line-edge variation that hit yield and device variability. The Intel 22 nm, 14 nm, and 10 nm fin cross-sections show how FinFET nodes have improved over time. 22 nm fins were a first-generation structure, relatively short, wide and strongly tapered. The shape limits current density and reduces gate control uniformity across the height of the fin. At 14 nm and 10 nm, Intel pushed the fins taller and narrower while also making the sidewalls more vertical. Rather than shrinking the device, these changes increase the effective channel width per fin and improve electrostatic control. The trade-off is that taller fins at tighter pitches are much more difficult to manufacture. Now, let’s compare the Helio G99 on TSMC N6 with the Kirin 9030 on SMIC N+3. Both processes are in the same class, with fin pitch of 30-32 nm on N+3 and 34 nm in our N6 cross-section. The pitch for N6 is especially interesting as N7’s HD library is generally listed with a 33 nm fin pitch, and N6 did not shrink pitches directly. Its density gains came from DTCO instead of tighter pitches. The 34 nm pitch was stable across our sampled region and serves more as a comparison against the SMIC N+3 we have not investigated further. Pinning down N+3’s fin patterning scheme takes more than one core unit. The CPU cores show a dense ~32 nm pitch, with the pitch between N-P fin pairs alternating between 78 and 88 nm. Logic alone may be consistent with dual-pitch mandrels of 120 and 110 nm, but this is a complex and unusual approach. Combining the pitch from the the 8T SRAM, which has more complex repeat unit, with the CPU core sequence allows us to reverse engineer the patterning steps with more confidence. As both the logic and SRAM should share the same base grid, a single CD mandrel lithography pattern with 128 nm pitch undergoing SAQP produces a die-wide ~32 nm grid (128 nm/4), which supports the pitch sequencing seen in both logic and SRAM cells. In the sampled cross-sections, N+3 shows a taller, narrower, higher-aspect-ratio fin than N6. The measured fin aspect ratio is ~9.5:1 on N+3 versus 7.8:1 on N6. N+3 also shows less top rounding, with an estimated radius of ~2 nm, compared with 2.8 nm on N6. Even though the fin widths differ, the ratio of top rounding to fin width tells the same story, with N+3 at 0.37 and N6 at 0.44. In a geometric sense, lower is better; a perfectly rectangular fin would have no top-rounding penalty. These are single-digit-nanometer features measured from a handful of cuts, so treat the absolute numbers as approximate. The important result is the relative gap: N+3’s fins are consistently taller, narrower and less rounded than N6’s. We’re diving deep into the most advanced datacenter and AI hardware hitting the market. To learn more about what’s in the pipeline or to commission a custom teardown, contact sales@semianalysis.com. Interested in joining us on this ride and think you can be a difference maker? Check out our Careers page. Standard Cell A standard cell is the basic building block of chip layout: a fixed-height row pairing one NMOS and one PMOS transistor that share a gate, tiled in a grid to build logic blocks. The key dimensions are contacted gate pitch (CGP), cell height (CH), fin count, and the lower-metal routing grid. To measure density, we use the Bohr metric: a weighted average of NAND2 gate area (60%) and scan flip-flop area (40%). This represents a realistic mix of combinational and sequential logic. This metric has its limitations, especially for complex cell layouts like TSMC’s FinFLEX, which alternates cells with different fin counts. Even so, it is the best metric for a pure process-level comparison. Another important measurement is the fin pitch; it refers to the distance between two fins of the same transistor. In a FinFET process, multiple fins are used in each transistor to increase the drive current and thus performance. TSMC N6 ships both a high-density (HD) library with 2 PMOS and 2 NMOS fins per cell, and a high-performance (HP) library with 3 of each. More fins under the shared gate mean more effective channel width. HP cells switch harder at the cost of area. Designers mix the two on a die, primarily spending HP cells on timing-critical paths, and matching their PPA targets. In the Cortex-A55 core of the Helio G99, we found a cell height of 240 nm for the HD cell. MediaTek has used HD cells in the G99 to minimize die size and thus cost. As an SoC for budget smartphones at ~$100, this is essential. By contrast, we found only one library in the Kirin 9030, with 2 NMOS and 2 PMOS fins. This suggests a narrower library strategy than TSMC N6, where both HD and HP libraries are widely used. This likely reflects the smaller customer base and the more constrained domestic design and electronic design automation (EDA) ecosystem. In all three CPU cores of the Kirin 9030, we found cell heights of 228 nm, 5% smaller than on N6. This is also a reduction of 9.5% over SMIC N+2’s cell height of 252 nm. SMIC N+3 and TSMC N6’s HD library both feature a CGP of 57 nm. For SMIC, this is a 9.5% shrink over N+2. In the past, CGP and cell height alone may have been enough to compare transistor density. Now, however, we must consider scaling boosters and DTCO as well. SMIC’s density gain does not come from EUV. It comes from using every available DTCO booster aggressively. First is fin depopulation: reducing the number of NMOS and PMOS fins in each cell. The first FinFET nodes started with 3 or 4 fins for each transistor. SMIC N+3 and TSMC N6 HD both use only 2 fins per transistor, trading drive strength for density. Next is contact over active gate (COAG). By landing the gate contact directly over the active gate, instead of out over the isolation region, the cell height drops. N+3 integrates COAG while N6 does not. Our N+3 gate-cut cross-sections indicate COAG, with the gate contact sitting over the active region, while N6 shows an off-gate contact. Last is single diffusion break (SDB). Diffusion breaks are inserted between cells in the same row to provide electrical isolation, but they also introduce local layout effects (LLE), layout-dependent shifts in electrical characteristics. In the past, a double diffusion break was used, consuming the space of two CGPs. SMIC N+3 and TSMC N6 instead use SDB, saving area but increasing LLE sensitivity. This must be controlled at the process level and accurately modeled in the process design kit (PDK) so that EDA tools can account for it. Overall, SMIC N+3 has a transistor density of 113.4 MTr/mm², slightly above TSMC N6 at 107.7 MTr/mm². Even without EUV, SMIC has achieved density beyond TSMC’s mature N6 node which utilizes EUV. Metal Stack The smallest critical dimension in the teardown is M0; SMIC N+3 uses a 32.5 nm local metal pitch. That is smaller than the 36 nm M0 pitch on Intel 18A in Panther Lake. However, this does not mean that SMIC has a better process than Intel 18A or TSMC N3P. M0 is a local intra-cell routing layer. Its usefulness depends on the full interconnect stack: M1 and M2 pitch, track count, via and line resistance, design rules, mask count, overlay control, and routing flexibility. The 32.5 nm M0 is consistent with self-aligned quadruple patterning (SAQP), whose four-population line-width loading we read coarsely as alternating widths of 21.5 to 24 nm; M1 and M2, at 38 and 40 nm, are consistent with self-aligned double patterning (SADP), a single A/B split. On TSMC N6, M0, M2, and M3 sit at a relaxed ~40 nm and are consistent with SADP-class double patterning, with no need for quadruple patterning. That said, we measure M2 for example at ~43 nm, likely inflated by sparse routing. We do not assign any specific layer to EUV from our cross-sections; the distinction we can draw is double versus quadruple patterning, not lithography wavelength. Transistor-level density in the front-end-of-line (FEOL) sets an upper bound, but the design is ultimately limited by what the interconnect stack can route. The lowest metals are the most important for standard-cell density, but the semi-global and global layers determine how usable that density is at the block and chip level. Two axes are commonly used for chip cross-sections: the fin-cut and the gate-cut. The micrograph above is a fin-cut and shows metals 0 through 3. This axis lets us see the even-numbered metals, with M0 right above the fins. There are two kinds of M0 lines. The first are the power rails; these are wide wires for the VDD and VSS running horizontally at the top and bottom edges of each standard cell. The wide wires measure 55 nm across, more than double the other M0 lines. Their width minimizes resistance and reduces IR drop. The second kind are intra-cell wires, short segments within the cell that connect terminals to M1. These have alternating widths between 21.5 and 24 nm. The M0 pitch is 32.5 nm, a 19% reduction versus N+2 and N6. At this pitch, DUV patterning requires more aggressive multi-patterning, increasing mask count, overlay sensitivity, process complexity and cost. M0 is below what a single DUV-defined spacer (SADP) can resolve, so SMIC cascades a second spacer step (SAQP). The cross-section reflects the cost: the M0 trenches are visibly more re-entrant (narrower at the bottom than the top) than M1 or M2 on the same chip and carry a bright barrier-rich foot where the trench meets the etch-stop layer. That shape is partly the intended damascene profile, as a slightly narrow bottom helps void-free copper fill, but its magnitude at M0 is driven by the tight pitch and the higher trench aspect ratio. Intel 18A supports an M0 pitch of 32 nm, although Panther Lake has only shipped with a looser 36 nm pitch. This is due to Intel’s heavy usage of HP libraries. Among leading-edge nodes, 18A has the loosest M0 pitches due to PowerVia. With power routing moving to the backside, congestion is reduced, and the entire front-side metal stack can be used for signal routing. M2 is the first true inter-cell routing layer. It runs horizontally like M0 but spans across multiple cells to carry block-level signals. The M2 pitch sets the cell’s track height – the number of M2 tracks that fit between the VDD and VSS rails, defining what the library calls a 6-track or 7.5-track cell. This layer is the most important, limiting the routing of entire blocks. SMIC N+3 features a 5.7-track cell. The M2 pitch is 40 nm, a 5% decrease over N+2 and the same as N6. This shrink keeps the pitch at the edge of what is possible with double patterning. Future nodes will need to increase the number of masks for M2 as reducing the number of tracks is much harder due to the limitations in routing. The micrograph above is in the perpendicular direction, the gate-cut, and shows metals 0 through 4. This allows us to see and measure the odd-numbered vertical metal layers. The M1 pitch is 38 nm, 9.5% less than N+2 and 33% less than N6. The M1-to-gate ratio matters because it sets local routing flexibility. N+2 and N+3 use a 3:2 ratio, while N6 uses a 1:1 ratio, explaining the huge differences in M1 pitch. The more M1 lines there are compared with the gates, the more flexibility there is for power and signal crossing within the cell. Routing flexibility enables more complex and better cells. Clean fractional ratios are also preferred as a grid is periodic and improves layouts. The 3:2 ratio gives SMIC more local routing flexibility than a strict 1:1 grid, but it also complicates layout and patterning. This is a DTCO choice, with SMIC increasing process complexity to recover density and routability without EUV. This 3:2 ratio is not very popular in the leading-edge nodes. TSMC has only used it on N7+, the N5 family, and the short-lived N3(B). They have switched back to a 1:1 ratio for N3E. Intel has only used it on the 10 nm/Intel 7 family, with Intel 4, 3 and 18A all using a 1:1 ratio. Samsung is the only one still using a 3:2 ratio at the leading edge, using it in the SF4 and SF3 families. It remains to be seen if SMIC will remain at a 3:2 ratio or move to a 1:1 ratio with its future nodes. The industry is still actively exploring these local-routing ratios. At VLSI 2026, imec will be presenting work on even higher ratios, including a 2:1 scheme that can reduce area by up to 14%. We will be covering the conference in a future newsletter article. Subscribe We’re diving deep into the most advanced datacenter and AI hardware hitting the market. To learn more about what’s in the pipeline or to commission a custom teardown, contact sales@semianalysis.com. Interested in joining us on this ride and think you can be a difference maker? Check out our Careers page. The final local interconnect layer for N+3 is M3, with a pitch of 44 nm. The M3 pitch is the same as on N+2 and 10% larger than on N6. The semi-global layers carry the majority of block-level signal routing. They have a coarser pitch than the lower local layers. On leading-edge nodes, they are designed to sit at the limit of DUV single patterning. M4 through M11 pitches were found divided between 80–82 nm (M4–M6), 128 nm (M7–M10), and 148 nm (M11). Given limited sampling, it is possible these are divided further in dense routing areas. At the top are two giant metal layers, M12 and M13. These have kept the same pitches as N+2 at 1920 nm and 4600 nm respectively. While the lower layers’ pitches are generally fixed by the process and library, the upper layers vary much more in pitch and count, depending on the design. Even two smartphone SoCs on the same process can have wildly different metal stacks. The Helio G99 carries fewer routing layers, reaching coarse metal pitches of 850 nm by M9, while the larger and higher-performance Kirin 9030 keeps fine pitches until M11. SRAM At the leading edge, SRAM is much more difficult to scale than logic. TSMC’s latest nodes have seen little to no bitcell scaling, while logic still has more DTCO levers to pull. While looking for other logic libraries in the GPU compute units, we stumbled upon the SRAM. The most common type of SRAM has 6 transistors (6T), but this cell had 8 transistors (8T) instead. 8T SRAM adds two transistors to form a dedicated read port. Unlike a 6T cell, where reading disturbs the storage, the decoupled read port removes read-disturb, improving read stability and letting the cell be pushed harder for performance. At first glance, the cut looked like an unusual logic library, with each cell row having 3 fins of one polarity and 5 fins of another. The rows also alternated in orientation. Energy-dispersive X-ray spectroscopy (EDS) resolved our confusion. The cut had not landed on the GPU logic, but on the SRAM macro beside it. The unusual fin pattern was due to the SRAM library. We return to EDS in the process flow analysis behind the paywall. SRAM libraries are not like traditional logic libraries. Due to the unequal number of PMOS and NMOS transistors, they require specialized rules and layout libraries. They do not need the flexibility of logic libraries, so they are hyper-optimized for one purpose: dense, reliable memory. The SRAM cell we found is a 1:2:2-2:2 cell. This means there is 1 fin per pull-up (PU) PMOS transistor, and 2 fins per pull-down (PD) and pass-gate (PG) NMOS transistor. These 2 PU, 2 PD and 2 PG transistors would usually form a single 6T high-current cell (HCC). An 8T HCC adds a read-pull-down (RPD) and a read-pass-gate (RPG) NMOS transistor, each with two fins. We measured a cell height of 406 nm, which brings the bitcell size to 0.0463 µm². That is a theoretical peak density of 21.6 Mib/mm². We estimate that a 6T HCC would have a cell height of 292 nm and a size of 0.0337 µm². This is ~12% larger than a 6T HCC on Intel 3 and 4. We also estimate the 6T high-density cell (HDC) to have a cell height of 228 nm and a size of 0.0260 µm². This is coincidentally the same as the logic standard-cell height measured earlier. The estimate puts the cell near Samsung 7LPP/5LPP and slightly below TSMC N7/N6. That is a theoretical peak density of 38.5 Mib/mm². 6T HDC is arguably the most important cell as it is used for the largest caches in a chip, the L3 caches and system-level cache (SLC). Both the Kirin 9020 and 9030 have split the SLC into 4 banks to raise total SLC bandwidth. In the Kirin 9030, the SLC increased from 2 MiB to 3 MiB per bank. Correspondingly, the number of arrays within the bank also increased by 50%, from 16 to 24. Each array can store 128 KiB and forms an orderly pattern on the die shot. From the Kirin 9020 to the Kirin 9030, the area of a 128 KiB SLC array decreased from 0.0477 mm² to 0.0392 mm², an 18% shrink. The achieved density is 25.5 Mib/mm², 66% of the theoretical maximum. While the SLC was quite similar across both chips, the L3 has seen some major changes, particularly in terms of its layout. The total capacity also went up from 10 MiB to 12 MiB. Much like the SLC, the L3 is also split into 4 banks. In the Kirin 9020, an L3 bank consisted of 16× 128 KiB arrays and 16× 32 KiB arrays. However, an L3 bank in the Kirin 9030 instead consists of 48× 64 KiB arrays. In the Kirin 9020 L3, a 128 KiB array was 0.0513 mm² and a 32 KiB array was 0.0154 mm². The size of the 128 KiB array is different on the L3 and SLC as the assist circuitry for the two arrays differs depending on their purpose. In the Kirin 9030 L3, a 64 KiB array is 0.0210 mm². Although not a like-for-like comparison, normalized for capacity, it is 18% smaller than the 9020’s 128 KiB L3 array and 31% smaller than its 32 KiB L3 array. The achieved density is slightly lower than the SLC, at 23.8 Mib/mm², 62% of the theoretical maximum. Unlike the L3 and SLC, the prime cores’ private L2 cache uses a 2-bank design. As the prime cores’ L2 is latency-critical, it likely uses 6T HCC instead of 6T HDC. The 9020 has 16 arrays in each bank while the 9030 has 32. Each array has a capacity of 32 KiB. A 32 KiB array in the L2 shrank from 0.0171 mm² to 0.0142 mm², ~17% smaller. The density is 17.6 Mib/mm², ~59% of the theoretical maximum for 6T HCC. SRAM scaled well from N+2 to N+3, shrinking by ~19%, close to the theoretical logic shrink. The caveat is that N+2’s bitcells were unusually large, bigger than comparable 7 nm-class nodes, so part of the gain is catch-up rather than true scaling. With the insights from STEEL’s teardowns, we will be doing a deep dive into SRAM in a future newsletter article. Subscribe Everything above came out of a single STEEL teardown: die annotation, block-level area analysis and TEM cross-sections through logic and SRAM. We’re diving deep into the most advanced datacenter and AI hardware hitting the market. To learn more about what’s in the pipeline or to commission a custom teardown, contact sales@semianalysis.com. Future Roadmap The same cross-sections that pin down N+3 also show where SMIC can go next. Although N+3 is already close to the practical limits of DUV multi-patterning in several layers, SMIC still has a few scaling levers left. A theoretical N+4 would likely start with cell height. N+3 uses 5 M0 tracks between its power rails. Moving to 4 M0 tracks, as on SMIC N+2 and TSMC N6, could reduce cell height by roughly 15%. The routing grid is only one side of the shrink; the front end also must fit into the smaller cell. One possible FEOL lever is reducing the p-to-n isolation spacing from two diffusion grid units to one. Intel used this scaling booster on Intel 4, and TSMC did so on its N3 family. This path trades layout flexibility for density. Fewer M0 tracks reduce local routing resources, while tighter p-to-n spacing raises integration and design-rule difficulty. M2 is also constrained by the cell height shrink. For SMIC to maintain a ~5.7-track cell, M2 would need to move toward ~35 nm. That would move another layer into SAQP territory. SMIC could also reduce the CGP from 57 nm to 54 nm. Intel reached a similar CGP on Intel 10 nm/Intel 7 without EUV. The local interconnect is also tougher. If SMIC keeps the 3:2 M1-to-gate ratio, M1 would need to shrink to 36 nm and would likely require SAQP as well. If SMIC moves to a 1:1 ratio, M1 could relax to 54 nm, but it would give up routing flexibility. Under this theoretical path, we estimate that SMIC N+4 could reach a cell height of 198 nm and a CGP of 54 nm, implying a Bohr density of 137.8 MTr/mm², on par with TSMC N5 or Samsung SF4. However, the difficulty is cumulative. Each step is individually plausible, but together they make N+4 harder than the transition from N+2 to N+3. It will likely take longer, cost more, and carry less process margin. A theoretical N+5 would require a larger integration shift. One possible path is backside contacts (BSCon), moving power routing and source/drain contacts to the backside, which would reduce front-side routing pressure and enable another reduction in cell height. The front-side metal pitches could relax to reduce the process complexity. M0 would likely relax slightly to ~34 nm, and M2 and M4 pitches could relax further. CGP is unlikely to shrink much further. Even with EUV, 48 nm has been the practical limit for yield and process control. This approach would allow N+5’s cell height to fall to 170 nm and its CGP to 53 nm. This implies a Bohr density of 163.6 MTr/mm², on par with Intel 18A’s HP library. However, this would not make N+5 cost-competitive with the leading edge. It would reach a similar density through a much more expensive route. The integration difficulty rises sharply, with new process flows for backside alignment, wafer thinning, contact reveal, and backside metallization. Past this point, standard density and interconnect scaling become increasingly unattractive. That is where Huawei’s roadmap stops looking like a normal foundry roadmap and starts looking like a packaging roadmap. At ISCAS 2026, Huawei unveiled its tau (τ) scaling law, reframing process scaling in the time domain. τ is the time cost of data movement and processing: switching delays in transistors, RC signal propagation delays in circuits, compute, memory, and networking latency. Outside Huawei’s terminology, this is called system-technology co-optimization. This is Huawei’s answer to its lack of EUV lithography. Without EUV, planar density cannot keep pace with TSMC, Intel, or Samsung. If transistor density cannot shrink further, Huawei’s alternative is to shorten wires, reduce buffering, and stack logic vertically. “LogicFolding”, Huawei’s implementation of this new scaling idea, is, in practice, an aggressive 3D stacking approach. AMD V-Cache places SRAM above or below a CPU die. AMD’s MI350X places active interposer dies (AIDs) underneath accelerator and compute dies (XCDs), with the AIDs handling cache, IO interfaces, the network-on-chip (NoC) and embedded metal-insulator-metal (MIM) capacitors. With LogicFolding, parts of the same logic block are split across multiple active dies bonded face-to-face at ultra-fine pitch. This allows Huawei to shorten some critical paths and reduce buffer overhead, not merely add cache capacity or offload the IO and interconnect. Shortening wires is where the higher clocks come from. A large share of a modern core’s delay and energy budget goes into driving long interconnects and the repeater buffers along them. LogicFolding distributes a block’s critical-path gates across multiple stacked tiers bonded at a very fine pitch, so the bond interface behaves like an additional metal layer and the longest paths get shorter. That is how Huawei expects to recover frequency and efficiency it cannot get from the process alone. Huawei’s roadmap shows its intent. Prime core frequency is targeted to rise from 2.75 GHz in the Kirin 9030 to roughly 5 GHz by 2031, far beyond what planar scaling alone could deliver. Prime cores with 3.1 and 3.39 GHz clocks are being tested in its labs, although their power consumption is unknown. Beyond that, chips are in the design, simulation or pathfinding phase, meaning the frequencies are targets. However, the direction matters more: LogicFolding also helps with performance, not just density. The catch is that Huawei’s density claim is not directly comparable to foundry densities. A stacked design can report more transistors per package footprint by adding active layers, even if each patterned die remains well behind TSMC or Intel in front-end density. This is how Huawei can claim to reach foundry 14A-equivalent density by 2031. This is not a like-for-like foundry comparison, with Huawei using stacked logic and measuring density per package footprint. On a normalized Bohr-density basis, SMIC N+3 is ~114 MTr/mm², 38% less than Intel 18A’s HD library. Huawei’s 3D roadmap closes the gap by stacking active logic, reaching 215 MTr/mm² by 2030. In 2031, the roadmap density jumps to 295 MTr/mm², implying either a third active layer, partial EUV insertion or aggressive planar DUV scaling. Huawei’s methodology makes other foundries look much denser as well. Applying it to AMD’s MI450X with an N2 top die on an N3P base die yields a theoretical density of 460.2 MTr/mm² in 2026, compared with Huawei’s 295 MTr/mm² in 2031. This Kirin 9030 does not use LogicFolding, remaining in a conventional mobile SoC package. Instead, it forms the baseline for how far Huawei and SMIC can push planar scaling. Future teardowns of Kirin and Ascend chips will show both planar logic density and Huawei’s hybrid bonding solutions. Export controls changed China’s optimization problem rather than ending it. EUV restrictions raised the cost and complexity of leading-edge manufacturing without freezing it. SMIC reaches N6-class logic density through DUV immersion, SAQP and DTCO, while Huawei shifts more of the burden onto architecture, packaging and system-level integration. Future nodes will be tougher. N+3 still had room to tighten local metals and reduce cell height and CGP. Further scaling without EUV leaves fewer levers. More aggressive multi-patterning adds masks and overlay error. SMIC can keep pushing DUV, but each step will get more expensive and less forgiving. The design side is just as critical. Huawei had domestic EDA tools and flows before the Kirin 9030, with the Kirin 9000s, 9010 and 9020 making that clear. Huawei was able to ship multiple consumer SoCs on SMIC N+2 and N+3 while cut off from the Western EDA stack. US export controls restricted EDA tools for advanced chips in 2022 but did not target tools for more mature chips. In 2025, the US government briefly placed much broader restrictions on EDA software from Synopsys, Cadence, and others, before lifting them less than two months later as part of a trade deal tied to rare earths. Huawei has been unable to access those tools because it remains on a US trade blacklist. That forced Huawei, SMIC and Chinese academic institutions to build their own tools and flows. Researchers at Peking University recently announced a prototype EDA tool for Huawei’s LogicFolding architecture, which requires a new flow to handle the multilayer layout and floorplan. This is not the same as replacing the full Synopsys or Cadence stack, but it shows where domestic EDA is headed: toward tighter co-optimization between architecture, process and packaging. These advances are also diffusing into the Chinese ecosystem. SMIC is licensing its N+2 and N+3 processes to HLMC/Hua Hong at the government’s direction rather than by choice. If the same process learning feeds into Ascend accelerators for AI training and inference, the choke point shifts from one named fab to an ecosystem. Alibaba’s T-Head silicon arm and Cambricon, a Chinese AI chip designer that is expected to supply ByteDance, could also be major beneficiaries. Sanctions aimed at SMIC alone become less effective once the manufacturing knowledge has spread to other fabs and design houses. China is not closing the gap with Intel, Samsung and TSMC. The teardown shows the opposite in several places: no EUV, no backside power, higher process complexity, and visible trade-offs. But China is still advancing. If domestic chips become good enough for phones, inference, networking and security-sensitive workloads, they can matter strategically without matching TSMC at the leading edge. Behind the paywall, we show what else STEEL can do, with material and process flow analysis of SMIC N+3, and analysis of the Kirin 9030 package. We’re diving deep into the most advanced datacenter and AI hardware hitting the market. To learn more about what’s in the pipeline, access the full Kirin 9030 and SMIC N+3 analysis or to commission a custom teardown, contact sales@semianalysis.com.
Synthesia HR chief: Tokenmaxxing isn't the goal. Getting results is.
Companies are pushing workers to use AI. Synthesia's Laura Gonzalez says measuring token usage can reward activity over outcomes.
Taiwan’s tallest tree found with help of citizen science
Deep in Taiwan’s misty mountains, researchers have confirmed the tallest tree in the country: a thousand-year-old fir tree higher than a 20-story building, which they’ve named “the heaven sword of the Da’an River.” Climbers scaled the tree and dropped a measuring tape from the top to the forest floor during the Lunar New Year holiday […]
Why did measuring Earth’s true shape matter for science?
The extraordinary challenge of determining the true shape of Earth reveals the deep value of measurement to scientific progress
How We Ranked the Best Companies for the Future
What we are measuring, why we are measuring it, and where we might be off base.
Oura Ring 5 Unveiled With Smaller Design and Blood Pressure Tracking
Oura today unveiled the Oura Ring 5, a significantly smaller smart ring with new health-monitoring capabilities including blood pressure trend detection, nighttime breathing analysis, and tools for GLP-1 medication tracking. The Oura Ring 5 is 40% smaller than its predecessor, measuring 6.09mm wide and 2.29mm thick compared to 7.99mm and 2.88mm on the previous model.
Pirated Sports Streams Are Warping TV’s Most Important Ratings
Advertisers love live sports. But good luck measuring that investment. Although tides of ad revenue flow based on the ratings of certain tentpole TV events, a new crop of scammers now operate illicit sports livestreaming rings, and there’s almost nothing broadcasters can do about it. According to a new report released Tuesday by ad tech […]
Scientists discover massive natural hydrogen source beneath Canada
Scientists in Canada have discovered that ancient underground rocks are naturally producing hydrogen gas — and lots of it. Measurements from mine boreholes in Ontario show the gas can flow continuously for years, offering a potential new source of clean energy called “white hydrogen.” Researchers say this hidden resource could help power industries and remote communities while cutting carbon emissions and reducing dependence on fossil fuels.
Scientists Are Counting The Diseases Climate Change Causes
A new report from the American Academy of Microbiology examines how attribution science is measuring the infectious disease burden caused by climate change.
Scientists discover hidden math secret inside Chinese money plant leaves
Scientists have uncovered a hidden mathematical secret inside the leaves of the Chinese money plant: a naturally occurring geometric pattern known as a Voronoi diagram, something typically associated with city planning, computer science, and network design. By mapping tiny pores and looping veins in the plant’s leaves, researchers discovered that the plant organizes itself using the same kind of elegant spatial logic humans use to solve complex distance problems — without ever “measuring” anything.
Quantum breakthrough could revolutionize teleportation and computing
Scientists in Japan have developed a new way to instantly detect elusive quantum “W states,” a major milestone for quantum technology. The breakthrough could help unlock faster quantum communication, teleportation, and powerful new computing systems.
LHU student revolutionizing stream restoration with rock-powered science
Natalie Brown kneels beside a humming wet lab setup at Commonwealth University-Lock Haven, calipers in hand, measuring the promise of Pennsylvania’s streams. The master’s candidate in biology regularly tests how everyday rocks could supercharge freshwater mussel recovery in the West Branch Susquehanna River, blending her geology roots with cutting-edge experiments. Her work is directly supporting […]
A one-in-a-million supernova seen five times could reveal the Universe’s true speed
A spectacular cosmic event nicknamed “SN Winny” could help solve one of astronomy’s biggest mysteries: how fast the universe is expanding. This rare superluminous supernova, located 10 billion light-years away, appears five times in the sky thanks to gravitational lensing, creating a dazzling “cosmic fireworks” effect. By measuring the slight delays between each appearance—caused by light taking different paths around two foreground galaxies—scientists can directly calculate the universe’s expansion rate.
How science can create a more consistent cup of coffee
Scientists borrowed technology used in battery research to show that measuring the current that runs through coffee can quantify a cup’s strength and flavor.
The Tech Elite Sure Have a Boring—and Expensive—Idea of Immortality
If you’re one of the lucky few who can afford to live forever, at least do something more interesting than measuring the “quality and durability of nocturnal erections.”
Patients taking weight-loss drugs often make 5 critical mistakes, doctor warns
Experts say GLP-1 drugs offer benefits beyond weight loss, but common mistakes like skipped labs and poor nutrition can undermine results or cause harm.
Doctor: Don't use fitness trackers, is obsessed with 1 health metric
Measuring our health data with wearable tech is more popular than ever. But a doctor said a lot of products make money from out health anxieties.
Young voters shun Trump, plan to vote blue in midterms, poll says
With the midterm primary season underway, a new national poll is measuring how the nation's youngest voters feel about Trump and the November election.
The Universe is expanding too fast and scientists still can’t explain it
A major international effort has produced an ultra-precise measurement of the Universe’s expansion rate, confirming it’s faster than early-Universe models predict. By linking multiple distance-measuring techniques, scientists ruled out simple errors as the cause of the discrepancy. The persistent “Hubble tension” now looks more real than ever. It could mean our current model of the cosmos is incomplete.
“AI polls” are fake polls
A few weeks after Donald Trump’s second presidential win, I took the train up from London (where I was living at the time) to Oxford to attend a conference on polls and forecasts of the 2024 election. Most of the attendees were pollsters or academics, but I also watched presentations from Aaru and Electric Twin, two companies that do what is interchangeably called synthetic sampling, silicon sampling, or synthetic audiences. Stripped of startup jargon, that means they use large language models (LLMs) to simulate responses to public opinion polls by having AI agents take on the role of survey respondents. I had already heard of Aaru thanks to some articles with eye-catching headlines like “No people, no problem: AI chatbots predict elections better than humans” in the months leading up to Election Day. The founders were making some big, some might even say far-fetched claims, such as: “within two years, we will simulate the entire globe — from the way crops are grown in Ukraine to how that impacts production of oil in Iraq, trade through the strait of Malacca, and elections for the mayor of Baltimore.” When Semafor asked Aaru’s cofounders — Cameron Fink and Ned Koh — about my boss, they said “we respect all those who came before us.” Nate (as he so often does) shared his thoughts on Twitter: Fink and Koh were relatively good-natured about this back-and-forth when we spoke at Oxford. They even offered to mail me one of the t-shirts featuring Nate’s quote they apparently had made. I never took them up on the offer, which I now somewhat regret. These synthetic sampling companies fell off my radar for a while, but they do still exist. In fact, Aaru recently received a $1 billion valuation. Is what they’re doing anywhere close to the most important frontier in AI development? Not by a longshot, especially when Anthropic just developed a model so adept at exploiting software vulnerabilities that it’s only being released to 40 companies. Still, silicon sampling is increasingly finding its way into public polling. Axios reported in March that “a majority of people trust their own doctors and nurses” based on findings from Aaru — without mentioning that the “people” in that sentence were actually LLMs. Around the same time, the Public Sentiment Institute “boosted” their online sample of 373 real survey respondents with 114 AI agents.1 (Spoiler alert: even the co-founder of Electric Twin doesn’t think that’s a particularly defensible approach.) Polling companies like Qualtrics and Ipsos are also developing synthetic data panels. So, what should we make of these … “polls”? Let’s get one thing out of the way: whatever they are, they’re not polls in the way that term is usually defined. Subscribe You can’t replace polls with AI On one hand, using LLMs to essentially make up fake survey respondents sort of sounds like the dumbest idea ever, one that will at best imperfectly replicate real polls while introducing all sorts of biases. On the other hand, with LLMs improving at a remarkable, perhaps even alarming rate, maybe that means I’m a dinosaur at the ripe old age of 24 because I still want to rely on polls that talk to actual people. I’m not going to argue that synthetic samples are completely useless. In fact, as I’ll return to later, there is evidence that some techniques can replicate topline survey results quickly and cheaply. But the marketing from certain companies can be slightly optimistic. “No traditional poll will exist by the time the next general election occurs,” said Fink in 2024. We’re just 206 days away from the midterms, and based on the fact that I still have to collect a bunch of polls every day, I’d say he should have run that prediction by a sample of AI agents before the interview.2 To see why synthetic samples can’t replace polls, here’s a quick primer on how they work. The simplest version of these models involves taking a LLM (like ChatGPT or Claude), giving it a demographic profile (e.g., a white, college-educated woman who lives in Utah and makes $70k a year), and then asking it to respond to a survey question. You repeat that process a few thousand times using different demographic profiles and end up with a sample of synthetic survey responses. The actual models used by private companies are more sophisticated than this, usually because they incorporate more hypothesized demographic characteristics for each agent and provide them with extra information. Aaru, for example, feeds agents a diet of news and information they’d be likely to consume, while Electric Twin incorporates their customers’ proprietary data about the audience they’re trying to replicate. The way Ben Warner, the co-founder of Electric Twin, explained it to me was “we have a large amount of data on […] for instance, 5,000 people. Can we make an accurate prediction of how they would respond to another question?” Still, it should be obvious why synthetic samples can’t replace polls. Polling is fundamentally a data collection process. We might use surveys to make predictions by feeding them into election forecasts, but the main purpose of a poll isn’t prediction, it’s gathering new data about what people think and how they feel. Silicon sampling, on the other hand, produces no new data. It’s simply a model: you input LLM training data, demographic prompts, and a bunch of other information, and it spits out a prediction for what a poll would say. We love models here, but models aren’t polls. That difference is an important philosophical sticking point for most pollsters I talk to. “I think politics should stay away from [synthetic sampling], because we’re trying to […] represent the voice of the people,” said Natalie Jackson, a vice president at GQR Insights. Democratic pollster John Hagner told me: “I think I’m just incredibly skeptical of this idea. I don’t think it’s research. At that point, you’re asking the machine to tell you what you already believe.” Hagner has seen some presentations of early synthetic sampling experiments, but so far, “if it’s being used in a campaign, people are keeping it incredibly quiet.”3 But Eli, I hear you saying, aren’t polls themselves increasingly governed by modeling decisions? Indeed they are: pollsters’ choices on which sampling method to use, how to define their likely voter models, and how to weight their samples can and do lead to dramatic differences in the results they publish. Aaru even referenced these limitations in the methodology statement included with that maternal mortality “poll” — although I’m using the term “methodology statement” loosely here, because it doesn’t really explain how the model works at all. We can ignore the (frankly preposterous) implication that synthetic sampling isn’t subject to a separate set of biases. The important point is that there’s still a meaningful difference between using weighting and other statistical techniques on actual polling data and using a model to predict what a poll would say. The latter is far closer to election forecasts or techniques like MRP — potentially useful models, but not a replacement for polls.4 To be fair, other synthetic sampling companies are perfectly happy with the distinction between polls and models. Warner compared polling and synthetic sampling to different tools in a toolbox. “The mistake I think we make is we think that these new tools should either work in exactly the same way or somehow replace these old tools,” he said. “Rather than thinking of it as, okay, so we’ve always had the hammer, we’ve always had the screwdriver, now we’ve got a saw. But don’t use a saw to try [to] do the job of a hammer.” A quick comment from Nate Eli didn’t ask me for a comment — rather rude of him, don’t you think? But since I’m editing this story, I figured I’d add a few quick thoughts rather than putting words in his mouth. Beyond the frequently misleading marketing, what bothers me about the AI “poll” hype is that as AI tools make statistical inference cheaper and/or better (note that these are not synonyms) that actually increases the comparative value of collecting original data. You might be able to train a model to make a reasonable estimate of what some hard-to-reach poll respondent would say — say, a young Black man who voted for Trump. (Such a person checks a number of boxes for a voter who is usually hard to reach in surveys.) Indeed, this is closely related to what models like the Silver Bulletin forecast already do. They essentially smooth out the kinks in noisy survey data by making inferences based on past voting patterns or national polls or surveys of other states. But you don’t actually know what these voters think unless you’re reaching them directly. If there’s a shift in opinion among this subgroup, you’re not going to detect it. So if I were running a campaign, I’d invest more in going the extra mile to find a representative sample of those voters. And then I’d hire some smart quants — or Claude? — to figure out the implications for campaign strategy based on proprietary data that my competitors didn’t have access to. -Nate Silver Are these models any good? If synthetic surveys are just a new type of model, the next obvious question is whether the models are at least accurate. The answer very much depends on who you ask. On one end of the spectrum, you have the maximalist argument that synthetic sampling is more accurate than actual polls. “It’s an incredibly challenging problem to go to someone and say ‘hey, we’re going to be more accurate at predicting human behavior than you, even when you talk to your customers directly’,” Koh recently told CNBC. In his view, synthetic sampling isn’t a saw to polling’s hammer, it’s “magic.” There’s certainly evidence that synthetic samples can replicate certain survey toplines. But if Aaru does have any examples of their approach outperforming the polls, they’re keeping those to themselves.5 Aaru’s 2024 election model, for example, had Kamala Harris leading in Michigan, Nevada, Pennsylvania, and Wisconsin on November 4th. And although they’ve since taken down their forecast page, they gave Harris a 50.5 percent chance of winning the race on November 2nd.6 After the election, Fink told Semafor he was happy enough with those results because they were “within margin of error,” a term that is completely meaningless when applied to a “sample” of AI agents. And of course, Aaru says their models have improved since 2024, so supposedly now they’d be more accurate than the polls? Still, their stronger argument is on cost: “We are significantly faster and cheaper than traditional polling, and still more accurate,” said Fink. The first two claims are undeniably true, but the third brings us to the opposite end of the spectrum. Both Jackson and Hagner are skeptical that these models are reliable for anything beyond replicating common survey toplines. “I just […] don’t think the machines are what we want when we’re looking for nuanced views. My example on this is people in Arizona and Nevada in 2024 who voted for Trump and voted for expanding abortion in their states on ballot initiatives,” said Jackson. Hagner identified another issue. Maybe the synthetic respondents, like sycophantic LLMs, are inhuman in one important way: they’re too nice. “The reports that have come through at the meetings that I’ve been at are that the early experiments on this, they cannot get respondents to be as racist or sexist or, frankly, as negative as human respondents,” he said. Academic research mostly agrees on this point. While there are some papers that show promising results when using LLMs to replicate polling data, most show that LLMs suffer from various quirks like producing too few “don’t know” responses and can seriously overpredict the favorability of politicians like Donald Trump and Kamala Harris. They also seem to struggle with too little variation between demographic subgroups, so the difference in predicted opinion between Democrats and Republicans, for example, is too small. When I asked Warner about these studies, his response to these papers was that just because academics can’t get synthetic sampling to work doesn’t mean that the technique doesn’t work in general. “Actually, the argument is, okay, yours does not [work]. That does not mean […] for this complex set of machinery, which uses a lot of investment, a lot of time, a lot of money, you can’t get it to work.” Cards on the table, I’m somewhat sympathetic to this argument because academics aren’t exactly great at making election forecasts. Usually, the people with skin in the game are the most accurate. Warner’s argument is that the approach Electric Twin takes — which includes, for example, making multiple predictions for each synthetic respondent using different models and prompts and subsequently averaging those to get a final prediction in a sort of ensemble forecast — produces better results than the simpler academic models. Warner shared a comparison between his method and the method from a recent academic paper with me, and Electric Twin was indeed able to get more accurate replication. But even still, he acknowledged that synthetic sampling “is not a crystal ball.” “If you asked me, do I think using other data sources will be more accurate than asking somebody who they will vote for, I would probably say no. But if you asked me ‘would your system be useful for our turnout modeling today?’ I would say yes.” For better or worse, it looks like the method is already getting more popular in the market research world. Most of the clients Aaru touts these days are businesses like EY and McDonald’s. And AI will probably begin to pop up in other parts of the political polling process. Pollsters are already using it to code open-ended survey responses, and some firms, like YouGov, are testing using LLMs to ask survey respondents questions. More worryingly, one danger to actual polls is that AI agents can be used to infiltrate online surveys. Most online polls use various checks to prevent that from happening, but there’s conflicting evidence on how effective those filters are and how prevalent AI agents currently are in online panels. If those agents ever become impossible to detect, it might spell the end of online polling, but the solution isn’t to replace all of your respondents with ChatGPT. Silver Bulletin is a reader-supported publication. To receive new posts and support our work, consider becoming a subscriber. Subscribe 1 That particular poll obviously doesn’t meet Silver Bulletin standards for aggregation. But we exclude all Public Sentiment Institute polls from our averages because we classify them as an amateur polling firm. 2 You could argue that Fink meant the next presidential election, but (a) I’m also confident we’ll still have real polls in 2028 and (b) in that case he should have asked an LLM to define “general election.” 3 Quick caveat: that’s reporting from a Democratic pollster. It’s possible that Republicans are more willing to use AI in political campaigns. 4 Indeed, Silver Bulletin does not include “polls” produced by MRP in our forecasts or averages, and we think it’s extremely misleading when their practitioners describe them in a way that suggests original data had been collected among a large number of states or Congressional districts. 5 A recent report from Aaru and EY did show two examples of a synthetic estimate being closer than a survey to a real-world benchmark — but I’d take those findings with a grain of salt because the report reads more like an ad and didn’t involve any sort of prediction being made ahead of time. 6 For comparison, our odds for Harris on the same day were 48.2 percent.
How a series of ‘left turns’ led to wearable health tech breakthroughs at UCLA
In an alternate world, UCLA professor Anne Andrews might be a CIA agent. But in this one, she’s a scientist breaking new ground measuring chemicals in the brain and body.
A global consensus establishes a basis for measuring and improving gut health
While “gut health” is becoming a popular term, mainly in media and marketing claims but also in the clinical context, researchers have not yet nailed down a definition. Now, a global group of 13 scientists and clinicians updates a consensus definition of gut health and implications from bench to bedside.
Gemma 4 and what makes an open model succeed
Having written a lot of model release blog posts, there’s something much harder about reviewing open models when they drop relative to closed models, especially in 2026. In recent years, there were so few open models, so when Llama 3 was released most people were still doing research on Llama 2 and super happy to get an update. When Qwen 3 was released, the Llama 4 fiasco had just gone down, and a whole research community was emerging to study RL on Qwen 2.5 — it was a no brainer to upgrade. Today, when an open model releases, it’s competing with Qwen 3.5, Kimi K2.5, GLM 5, MiniMax M2.5, GPT-OSS, Arcee Large, Nemotron 3, Olmo 3, and others. The space is populated, but still feels full of hidden opportunity. The potential of open models feels like a dark matter, a potential we know is huge, but few clear recipes and examples for how to unlock it are out there. Agentic AI, OpenClaw, and everything brewing in that space is going to spur mass experimentation in open models to complement the likes of Claude and Codex, not replace them. Especially with open models, the benchmarks at release are an extremely incomplete story. In some ways this is exciting, as new open models have a much higher variance and ability to surprise, but it also points at some structural reasons that make building businesses and great AI experiences around open models harder than the closed alternatives. When a new Claude Opus or GPT drops, spending a few hours with them in my agentic workflows is genuinely a good vibe test. For open models, putting them through this test is a category error. Something else to be said about open models in the era of agents is that they get out of the debate of integration, harnesses, and tools and let us see close to the ground on what exactly is the ability of just a model. Of course, we can’t test some things like search abilities without some tool, but being able to measure exactly the pace of progress of the model alone is a welcome simplification to a systematically opaque AI space. Share The list of factors I’d use to assess a new open-weight model I’m considering investing in includes: Model performance (and size) — how this model performs on benchmarks I care about and how it compares to other models of a similar size. Country of origin — some businesses care deeply about provenance, and if a model was built in China or not. Model license — if a model needs legal approval for use, uptake will be slower at mid-sized and large companies. Tooling at release — many models release with half-broken, or at least substantially slower, implementations in popular software like vLLM, Transformers, SGLANG, etc due to pushing the envelope of architectures or tools. Model fine-tunability — how easy or hard it is to modify the given model to your use-case when you actually try and use it. The core problem is that some of these are immediately available at release, e.g. general performance, license, origin, etc. but others such as tooling take day(s) to week(s) to stabilize, and others are open research questions — with no group systematically monitoring fine-tunability. In the early era of open models, the days of Llama 2 or 3 and Qwen pre v3.5, the architectures were fairly simple and the models tended to work out of the box. Some of this was due to the extremely hard work of the Llama, Qwen, Mistral, etc. developer teams. Some is due to the new models being genuinely harder to work with. When it comes to something like Qwen 3.5 or Nemotron 3, with hybrid models (either gated delta net or mamba layers), the tooling is very rough at release. Things you would expect to “just work” often don’t. I’ve been following this area closely since we released Olmo Hybrid with a similar architecture, and Qwen 3.5 is just starting to work well in the various open-source tools that need to all play nice together for RL research. That’s 1.5 months after the release date! This is just to start really investing more into understanding the behavior of the models. Of course, others started working on these models sooner by investing more engineering resources or relying on partially closed software. The fully open and distributed ecosystem takes a long time to get going on some new models. All of this is lead-in for the most important question for open models — how easy is it to adapt to specific use-cases? This is a different problem for different model sizes. Large MoE open-weight models may be used by entities like Cursor who need complex capabilities in their domain, e.g. Composer 2 trained on Kimi K2.5. Other applications can be built on much smaller models, such as Chroma’s Context-1 model for agentic search, built on GPT-OSS 20B. The question of “which models are fine-tunable” is largely background knowledge known by engineers across the industry. There should be a thriving research area here to support the open ecosystem model. The first step is to understand characteristics of different base and post-trained models to understand what they look like. The second step is to tune pretraining recipes for open models so they’re more flexible. Interconnects AI is a reader-supported publication. Consider becoming a subscriber. Subscribe For The ATOM Project and other Interconnects endeavors, we’ve put in substantial effort to measuring adoption trends in the open ecosystem. Everything takes a long time to unfold after a model is first publicly available — and adaptability is why. What we know for sure now, when Qwen has been going from strength to strength with its releases, is that technical staff across the industry has gotten comfortable working with Qwen models. Countless research methods and datasets were made to work with Qwen. It’ll take patience for any other model family to get to this point — a patience I’m not sure many open model builders have. This takes us to Gemma 4, Google’s latest open models. Gemma 3 was released more than a year ago, in March of 2025, and is a bit underrated. Gemma 4 comes in 4 sizes for now, with a bigger, MoE model of over 100B total parameters rumored but not released yet. The models we have today come in sizes of ~5B dense, 8B dense, 26B total 4B active MoE, and 31B dense. I’m most excited that they’re finally adopting a standard Apache 2.0 open source license. This’ll massively boost adoption. The standard of better licenses for strong open-weight LLMs was set by mostly Chinese open model labs in the last 1-2 years, and now U.S. companies are following suit. I will personally be so happy if the horrible Llama licenses and Gemma terms of service were an ~18-month transient dynamic of the industry being nervous about releasing strong open models. The Gemma 4 scores look very solid, the small models have incredible benchmark scores (especially in general domains like LMArena) and the 31B model rivals the recent Qwen 3.5 27B, which is the leading member of that class. The ~30B size range is an important one, as it’s accessible both to researchers and to enterprises looking to deploy the model in real use-cases. Where the 7B model scale is the default for tinkering and research, a 30B model is the default for seeing if an open model can unlock substantial value in your specific workflow — a good mix of intelligence, low price, tractability for downstream training, etc. This takes us back to the above adoption criteria I mentioned for open models and the bigger question — do I think Gemma 4 will be an overwhelming success? Previous Gemma models have been plagued by tooling issues and poorer performance when being finetuned. Gemma 4’s success is going to be entirely determined by ease of use, to a point where a 5-10% swing on benchmarks wouldn’t matter at all. It’s strong enough, small enough, with the right license, and from the U.S., so many companies are going to slot it in. I’m cautiously optimistic that Gemma 4 is going to work better here. Winds are shifting for open models built in America. We saw GPT-OSS go through a bumpy launch to become an overwhelming success. There’s a collective energy around the likes of Reflection, Arcee, Nemotron, Gemma, Olmo, and peers that show substantial demand for building new stacks around open models. There’s capital to be spent on AI stacks across the economy by those who want more ownership of everything, including the model. After launching The ATOM Project 240 days ago, the conversation is shifting into the next stage. Summer of 2025 was a crisis moment where the U.S. AI scene realized it can’t wait and figure out open models after building AGI. The two markets will capture different areas and proceed in parallel. Now that more companies in the U.S. are releasing strong models, we need to improve the ecosystem so that these models are easy to use, understand, and build value around. It’s the hard work to build another inflection point in these adoption plots I’ve been updating consistently, but that’s the work to be done. Join me in it. More data coming soon! Here’s a sneak peek:
Scientists Develop New Gut Health Measure That Tracks Disease
Scientists have identified a new way to distinguish healthy guts from diseased ones and track how some illnesses progress by measuring how gut bacteria interact with one another. According to a study published in Science, a Rutgers-led team of scientists found that healthy and diseased digestive systems behave like two distinct ecological states, driven not by individual microbes but […]
iPhone 18 Pro Max Thickness and Weight Allegedly Revealed
Apple's upcoming iPhone 18 Pro Max will be slightly thicker than its predecessor, measuring in at 8.8mm, up from 8.75mm on the iPhone 17 Pro Max. The information comes from oft-accurate Weibo-based leaker Ice Universe. The claim chimes with a report last year that alleged hardware changes in the iPhone 18 Pro Max will make it the heaviest iPhone yet.
This Overlooked Health Condition Could Raise Your Risk of Death by 83%
A study following more than 5,000 participants for 12 years found that straightforward clinical measurements can help identify sarcopenic obesity without relying on complex imaging methods such as MRI and CT scans. Researchers from the Federal University of São Carlos (UFSCar) in Brazil, working
NASA’s PACE Satellite Provides New Pollution Measuring Product
NASA’s Plankton, Aerosol, Cloud, ocean Ecosystem (PACE) satellite, known for measuring tiny organisms in the ocean and particles in the atmosphere, has a new
Simple blood test can forecast Alzheimer’s years before memory loss
Scientists have created a blood test that can estimate when Alzheimer’s symptoms are likely to begin. By measuring a protein called p-tau217, the model predicts symptom onset within roughly three to four years. The protein mirrors the silent buildup of amyloid and tau in the brain long before memory loss appears. This advance could speed up preventive drug trials and eventually guide personalized care.
Five Berkeley Lab Scientists Receive DOE Early Career Research Awards
Five scientists from Lawrence Berkeley National Laboratory (Berkeley Lab) have received Early Career Research Program (ECRP) awards from the Department of Energy. The ECRP program, now in its 16th year, was established to bolster the nation’s scientific workforce by supporting exceptional researchers at the outset of their careers, when many scientists do their most formative work. Recipients use the funding to pursue projects that have the potential to solve scientific challenges and advance the country’s STEM capabilities. To be eligible for the program, a researcher must be an untenured, tenure-track assistant or associate professor at a U.S. academic institution, or a full-time employee at a DOE national laboratory or Office of Science user facility, who is within 10 years of having earned their doctorate. Awards to an institution of higher education will be approximately $875,000 over five years, and awards to a DOE national laboratory or Office of Science user facility will be approximately $2,750,000 over five years. This year’s Berkeley Lab awardees and their projects are listed below: Timon Heim: Better detectors for more powerful particle colliders Timon Heim is a staff scientist in the Physics Division. His ECRP project will develop a new type of detector that can track the particles generated by collisions in particle colliders with high spatial and temporal resolution. Current detectors in use at high-energy colliders, like the ATLAS detector at the Large Hadron Collider (LHC) at CERN, can track the trajectory and measure momentum of particles, but they can’t track a particle’s time of arrival with sufficient precision to discern from which of the many particle interactions they originate. Heim’s proposed technology — called a 4D tracking detector, because it captures data about particles’ positions in space and time — aims to solve this challenge, allowing researchers to glean more precise data from collision experiments to uncover new insights into fundamental forces and potentially discover new particles. The enhanced capabilities of 4D detectors will be a necessity for experiments at future colliders, such as the Muon Collider or Future Circular Collider. The project will develop custom microelectronics that process the signal of a pixelated silicon sensor that can register charged particles passing through it in an extreme radiation environment that these colliders produce during operation. The goal is to measure individual particles’ position with 50-micro-meter spatial and 50 pico-second (one trillionth of a second) temporal precision, for an incoming flux of around 1 billion particles per second per square centimeter. Harrison Lisabeth: Studying underground processes to unleash abundant energy Harrison Lisabeth is a research scientist in the Energy Geoscience Division focused on combining real-world experiments and AI to guide development of geothermal energy systems. His project will explore how deep underground rock formations – where geothermal wells are dug – are stressed by chemical reactions of the minerals and mechanical strain from movements in the Earth. A better understanding of subsurface stresses is essential to engineer safe and reliable geothermal wells that access the heat of deep rocks to produce electricity. It is estimated that the U.S. could generate ten times the amount of energy needed to power the grid using geothermal sources. Lisabeth plans to simulate how minerals grow, dissolve, and influence the behavior of surrounding rock in geothermal reservoirs by tracking chemical changes in rock samples under pressure in the lab. The project will also develop machine learning models to predict geologic strain based on field measurements and refine existing models that describe how common rock types deform under stress. Combining these approaches, Lisabeth hopes to develop a framework to predict stress to improve stability and longevity of geothermal wells. Daniel Carney: Pushing the quantum boundaries of particle measurement Daniel Carney is a staff scientist in the Physics Division who uses quantum information science to study particle physics and gravity. At very small scales, “quantum noise” — random fluctuations from the act of measurement — can drown out faint signals from rare processes. Carney’s ECRP project, “Fundamental Physics at the Quantum Limits of Measurement,” aims to build a new mechanical quantum sensor that can measure extremely small particle interactions below the limits of that quantum noise. The approach uses lasers to hold a tiny bead containing radioactive atoms; when one decays, the system measures the bead’s motion and the energy of any particles emitted. The experiment’s initial goal is to search for heavy sterile neutrinos, hypothetical particles predicted by some extensions of the Standard Model. The system could also help measure the mass of the neutrino, study the way particles interact through the electromagnetic and weak forces, and hunt for low-mass dark matter — all unsolved mysteries in particle physics. Callum Wilkinson: Using machine learning to demystify neutrinos Callum Wilkinson is a staff scientist in the Physics Division and part of the international Deep Underground Neutrino Experiment (DUNE) hosted by Fermi National Accelerator Laboratory. DUNE will create an intense beam of neutrinos that interact in detectors in Illinois and, after an 800-mile journey through the earth, in massive detectors in South Dakota. By precisely measuring neutrinos and how they change, researchers will get a better understanding of their fundamental properties and whether they are responsible for why there is so much more matter than antimatter in our universe. Wilkinson’s ECRP project will apply two machine learning techniques to improve the analysis of DUNE’s data and enhance the experiment’s precision. One approach will look for hidden patterns in how neutrinos interact and produce low-energy particles, while the other will help correct for distortions in what the detectors record. The methods will be tested on one of DUNE’s prototype detectors and could be used in other neutrino experiments, improving their measurements and helping scientists build better models. Aditi Krishnapriyan: Machine learning methods that scale as datasets grow Aditi Krishnapriyan is a faculty scientist in Berkeley Lab’s Applied Mathematics and Computational Research Division (AMCR) and assistant professor at UC Berkeley. Her project, awarded through UC Berkeley, will develop machine learning methods that scale with dataset size for modeling complex, multi-scale phenomena. The resulting tools can be applied in a wide range of research, including the study of new electrolytes for batteries or the examination of molecular interactions. ### Lawrence Berkeley National Laboratory (Berkeley Lab) is committed to groundbreaking research focused on discovery science and solutions for abundant and reliable energy supplies. The lab’s expertise spans materials, chemistry, physics, biology, earth and environmental science, mathematics, and computing. Researchers from around the world rely on the lab’s world-class scientific facilities for their own pioneering research. Founded in 1931 on the belief that the biggest problems are best addressed by teams, Berkeley Lab and its scientists have been recognized with 17 Nobel Prizes. Berkeley Lab is a multiprogram national laboratory managed by the University of California for the U.S. Department of Energy’s Office of Science. DOE’s Office of Science is the single largest supporter of basic research in the physical sciences in the United States, and is working to address some of the most pressing challenges of our time. For more information, please visit energy.gov/science.
Tellus Science Museum opens interactive exhibit teaching children how to measure
Measuring might feel like math in the classroom, but at the Tellus Science Museum, it looks a lot more like playtime.
Citizen science and the challenge of measuring urban sustainability
Cities are expected to track sustainability progress with data that are often incomplete, outdated, or available only at national level. New research led by IIASA in collaboration with UN-Habitat finds ...
The Guardian view on microplastics research: questioning results is good for science, but has political consequences
Editorial: Errors in measuring microplastic pollution can be corrected. Public trust in science also needs to be shored up
Barton Health welcomes first baby of 2026
SOUTH LAKE TAHOE, Calif. – Barton Health welcomed the first baby of the New Year on January 6 at 2:36 p.m. Murphy Vera Ross, female, weighing 7 pounds, 11 ounces, and measuring 20 inches long, was born to Ashley Brier Ross and Barret Ross of South Lake Tahoe, CA. Dr. Megan Jewell, board-certified obstetrician and gynecologist...
Eurozone Business Confidence Slips
The decline in a key indicator measuring attitudes of businesses across multiple sectors ends four consecutive months of rising sentiment.
Longevity because a business worth billions for Silicon Valley
In the longevity industry, the goal isn't just measuring your body. It's reprogramming it to age more slowly
Withings thinks of its latest smart scale as a ‘longevity station’
Withings says the scale isn’t just for measuring weight.
Withings thinks of its latest smart scale as a ‘longevity station’
Withings says the scale isn’t just for measuring weight.
I Spent $160,000 of My Family's Savings to Bootstrap a Startup — Here's What No One Tells You About Funding
Every dollar counts, and you quickly learn to focus on paying customers and measuring ROI before chasing vanity metrics.
'Non-Political' Creators Have Outsized Influence on Audience's Politics, Study Finds
Measuring the Joe Rogan effect.