SCROLL NEWS / DISCOVERY
Search headlines
Find the stories shaping the conversation.
Results for " GPU " 379 found 🔀 AI-powered Shuffle
National Compute Grid Launches to Pool Idle AI Capacity Across Providers
For all the talk about a shortage of AI computing power, and the need for more powerful chips and bigger data centers, there’s another side to the story. A significant amount of existing capacity isn’t being used. While tech companies race to build massive new data centers, expensive GPUs in other facilities can sit idle […]
Broadcom Is Building an AI Business Unlike Nvidia’s. Here’s My Price Prediction
Broadcom competes with Nvidia without selling a single GPU to the general market, and that unusual strategy has locked in hyperscaler contracts stretching years into the future. Whether the stock has already priced in that advantage is the question worth answering.
Report: RTX 5090 GPUs discontinued by Nvidia, as GB202 is reserved for AI workstation cards
Not that I could afford one anyway.
Forty new games are coming to Xbox consoles next week.
Microsoft's gaming platform ecosystem is gearing up for a major expansion of its catalog. Over the next seven days, Xbox Series X|S, Xbox...
NVIDIA Reportedly Ending the RTX 5090, Reserves GB202 Silicon for Workstation GPUs
NVIDIA has reportedly stopped allocating GB202 silicon to the GeForce RTX 5090, according to leaker MEGAsizeGPU (@Zed__Wang) on X. "No more GB202 for 5090, confirmed," the leaker posted, adding in a follow-up that "all future GB202 supplies are RTX Pro exclusive." GB202 is the largest "Blackwell"...
NVIDIA CEO Jensen Huang receives National Medal of Science for turning gaming GPUs into AI powerhouses
NVIDIA CEO Jensen Huang received the National Medal of Science for advancing GPUs into general-purpose computing engines that power modern AI.
NVIDIA reportedly discontinuing RTX 5090, GB202 GPUs to be reserved for RTX PRO series
NVIDIA is rumored to discontinue its RTX 5090 lineup and introduce an RTX 5080 with 24GB. The claim comes from an unverified factory message.
NVIDIA-Backed Upscale Wants to Make Rival AI Chips Work Together
NVIDIA GPUs dominate today’s AI data centers, but they’re no longer the only processors companies want to use. Cloud providers are developing their own AI chips, while specialized accelerators are entering the market. Getting these different processors to work together efficiently is creating a new challenge for data center operators. That’s the problem Upscale hopes […]
Dell launches Nvidia RTX Spark laptop with 18-core CPU, 3.2K 120Hz OLED display and 99 Wh battery
Dell has launched its first XPS laptop powered by Nvidia’s RTX Spark platform. The XPS 16 combines an 18-core N1X chip, 5,120 GPU cores, 32 GB of unified memory and a 16-inch 3.2K Tandem OLED touchscreen.
Gears of War: E-Day GPU Benchmark: 44 GPUs Tested
We tested 44 GPUs in Gears of War: E-Day at 1080p, 1440p, and 4K to see what it takes to run the game well, how 8GB cards...
Grand Theft Auto VI developers literally forced Sony to develop QSSR technology
According to insider PainkillerQ8, the Rockstar Games team literally forced Sony developers to dig up the technology...
Valorant Player Banned Due To Previous GPU Owner’s Cheating
In a wild coincidence a German player was penalized over the playing habits of their Ryzen 7's previous owner, claiming that nothing short of replacing the entire PC will overcome Riot's infamous anti-cheat detection
The Amazing Work By Valve's Timur Kristóf On Improving Old AMD GPUs On Linux
Over the past year Timur Kristóf of Valve's Linux graphics driver team has made multiple very nice improvements to the AMDGPU kernel driver for enhancing support for old (GCN 1.0/1.1 era from a decade ago) graphics cards so that they can better handle Linux gaming and other tasks
Gears of War: E-Day Performance Benchmark Review
Gears of War: E-Day is a prequel set 14 years before the original game. Built on Unreal Engine 5 with RTX Mega Geometry, DLSS 4.5, FSR 4, XeSS, and ray tracing, we examine graphics quality, VRAM usage, upscaling options, and how it performs across a wide range of modern GPUs.
Borderless Gaming’s DLSS 5 Implementation Runs Without Touching NVIDIA’s Software, Quantized to Work Well Even on Unsupported GPUs
Borderless Gaming has been updated with a custom implementation of NVIDIA DLSS 5 Neural Rendering that runs on a much wider range of GPUs.
Borderless Gaming adds DLSS 5 Neural Rendering without modifying game files
Borderless Gaming can now apply DLSS 5 Neural Rendering without replacing game DLLs or modifying game files. The quantized BGFX effect can run on GPUs including the RTX 3090.
The Witcher 3 Remastered Crushes GPUs With Path Tracing, but Free Mod Claws Back Up to 30% Performance
Following its surprise announcement during this year's gamescom Opening Night Live, The Witcher 3 Remastered is now available on PC, PlayStation 5, Xbox Series X|S, and Nintendo Switch 2. While the number of improvements to gameplay is staggering and well thought out, what most were looking forward to was the visual uplift, which includes very demanding path tracing support on PC. Thankfully, a mod has already come to the rescue, as it did after the launch of the Next-Gen update, optimizing both ray tracing and path tracing for better performance with minimal loss of quality. "Faster RT/PT in The Witcher […]
The Witcher 3 Remastered showed 38 FPS with DLSS 5 versus 68 FPS for the Next-Gen version
Detailed performance and visual quality measurements of the updated version have appeared online. The Witcher 3: Wild Hunt Remastered using...
Not the GPUs. This Overlooked Piece of Nvidia's Business Could Matter Most for the Stock's Next Leg Up.
Nvidia sees this area as a $200 billion opportunity.
Not the GPUs. This Overlooked Piece of Nvidia's Business Could Matter Most for the Stock's Next Leg Up.
Nvidia sees this area as a $200 billion opportunity.
Modder Squeezes 74% More Performance Out of DLSS 5 on AMD Radeon in One Day
Modders working on bringing NVIDIA's DLSS 5 Neural Rendering to AMD Radeon GPUs just had their busiest day yet. The developer behind the DLSS-NR-on-AMD tool shipped two new versions within 24 hours, Alpha 0.3.3 and Alpha 0.4.0, together adding roughly 74% performance on Radeon hardware, according...
Amazon Just Knocked $900 Off Apple's Latest 16-Inch MacBook Pro
Amazon has introduced a major discount on the 16-inch MacBook Pro with the M5 Max chip, 32-Core GPU, 36GB RAM, and 2TB SSD, now available for $3,499.00, down from $4,399.00. This is a massive $900 markdown on the high-end MacBook Pro, and it's joining a few other MacBook Pro discounts on Amazon this week. Note: MacRumors is an affiliate partner with Amazon. When you click a link and make a purchase, we may receive a small payment, which helps us keep the site running.
This wireless eGPU claims to deliver RTX 5090-like performance over Wi-Fi
WiCi recently began inviting interested parties to sign up for updates about the WiCi One, which aims to become the first successful wireless external graphics card. The...
Your old GPU could be worth more than you think
Your old GPU could be a goldmine thanks to the RAMaggedon.
Silent Hill: Townfall compared on RTX 5080 and PS5 Pro with DLSS 4.5 and PSSR 2 in 4K
An extended comparative benchmarking of psychological horror films has been published online. Silent Hill: Townfall, in which the quality of graphics and level...
Sony may reconsider plans to phase out physical discs for PlayStation.
The management of Japanese company Nippon Ichi Software and its Western publishing division NIS America announced that they are holding direct consultations...
This External GPU Uses Wi-Fi To Transform Any Device Into A Gaming Rig
The WiCi One is available for preorders starting at $1,999.
Ace Combat 8 Wings of Theve requires at least 32GB of RAM to play comfortably.
We conducted detailed benchmarking of the Ace Combat 8: Wings of Theve flight simulator to measure the actual consumption of personal...
RTX 5080 and RX 7900 XTX showed almost identical FPS in Ace Combat 8 in 4K
We have carried out detailed benchmarking of the latest NVIDIA GeForce RTX series graphics accelerators and AMD Radeon RX in the flight simulator Ace Combat...
Ace Combat 8: Benchmarks of Graphics Cards and Processors
We ran extensive benchmarks of Ace Combat 8: Wings of Theve on graphics cards ranging from the RTX 3060 to the RTX 5090 and the RX 6750 XT to the RX 9070 XT in conjunction with CPU AMD and...
Intel Panther Lake Teardown, 18A, BSPD, GAAFET, SemiAnalysis STEEL
Panther Lake debuts the first commercial implementation of backside power delivery (BSPDN), introduces Intel’s first iteration of gate-all-around (GAA) transistors, and showcases their advanced packaging capabilities with its Foveros-S assembly. With Panther Lake, Intel’s manufacturing arc has shifted from nebulous roadmaps to shipped silicon, a significant milestone on their long road back to competitive semiconductor manufacturing. To evaluate the extent of Intel’s comeback, we tore down Panther Lake. The SemiAnalysis STEEL teardown lab breaks down advanced datacenter and AI hardware. To learn more about our pipeline or to commission a teardown, contact sales@semianalysis.com. WE’RE HIRING: Architecture, floorplan, packaging, manufacturing, and labs experts. Opportunities from system to transistor and everywhere in between. Check out our Careers page. Our teardown traces 18A from its four-sheet RibbonFETs (Intel’s marketing name for GAAFETs) and gate stacks through contacts, frontside and backside wiring, and the bonded carrier. We explain how these material and integration choices improve gate control and reduce resistance, while adding capacitance, thermal resistance, and process complexity. Our measurements put Panther Lake’s 18A compute logic and TSMC N3E GPU logic at similar logic density. However, 18A does not lead TSMC N3P, N2 or Samsung SF2 in peak density. Panther Lake’s CPU cores are incremental updates, and the high-end GPU still uses TSMC N3E. Panther Lake assembles one compute tile, one GPU tile, and one I/O tile atop a passive base tile using Intel’s Foveros-S advanced packaging. Both compute tile variants use Intel 18A. The Xe3 GPU options are a 4-core GT1 tile on Intel 3 and a larger 12-core GT2 tile on TSMC N3E. Both I/O tile variants use TSMC N6. [1], [2] Our analysis centers on the PTL-U compute tile, both the 4-core and 12-core GPU tiles, as well as the 12-lane I/O tile. In conventional chips, power and signal are routed through the same frontside metal stack towards the device frontend. Power rails consume scarce routing resources near the transistors, while tall via stacks carry VDD and VSS from the coarse upper wires to local rails. Backside power delivery (BSPD) moves the main power network behind the transistor layer, to the backside, separating it from frontside signal routing. We covered BSPD and its impacts in 2024. [3], [4], [5] Intel’s BSPD implementation, branded as “PowerVia”, routes power through dedicated backside metals to nano-TSVs, which connect those rails to local source/drain (S/D) contacts. Implementing that separation requires Intel to build the interconnect stacks from both sides of the wafer. The frontside comprises the M0-M14 signal stack, while the backside comprises the BM0-BM5 power stack. M0 and BM0 are closest to the transistors. The nano-TSVs connect the two sides, but Intel patterns and etches each via from the front after forming the contacts. A narrow via runs from the side of the contact deep into the silicon substrate. Intel then completes the frontside signal metal stack, bonds the wafer to a carrier, flips it and removes the original substrate until the buried via tips are exposed. The backside metal stack is then deposited directly on the revealed vias. The nano-TSV and backside-via profiles taper in opposite directions because Intel forms them from opposite sides of the wafer. The transistor structures form the FEOL. Local contacts and nano-TSVs connect them to the wiring. M0 begins the frontside interconnect stack. The silicon carrier remains attached above the frontside interconnects. It supports the device wafer during substrate removal and backside processing and remains part of the finished chip’s thermal path. PowerVia removes the main power distribution from the congested frontside metals, routing supply through shorter and wider backside wires. Its lateral landing still occupies area in the standard cell, so it recovers less cell area than a direct backside contact. [3] Nano-TSVs beside the logic devices carry VDD or VSS from the backside power network, while signal connections continue upward through the frontside metals. Backside Interconnects Samsung SF2 data is included for comparison to Panther Lake’s within this article. SF2 is the incumbent GAA foundry node but lacks BSPD, serving as a useful reference to evaluate 18A. A full teardown of Samsung’s S26 products, processed on SF2, will be shared soon.Nanosheet-cut EDS comparison. The PowerVia supply path runs from the backside Cu rails through Mo-lined W nano-TSVs to the local transistor contacts. In this cross section, the tapered connection spans roughly 150 nm from the contact level to BM0. The Ta liner confines Cu and promotes adhesion to the surrounding stack; the AlOₓ etch stop controls the next dielectric etch above the rail. Dielectric beneath the ribbons electrically separates the devices from the backside wiring and removes the conducting silicon body below the channel. [6] AlOₓ serves as an etchstop (ES), enabling endpointing and protecting the underlying layers. Low-volatility aluminum fluoride reaction products resist the fluorinated plasma, allowing a thin AlOₓ film to protect the metal while the surrounding low-k dielectric is removed. [6], [7]. While the BM0 and layers above the M1 lines show double AlOx layers, Our SMIC N+3 teardown showed single AlOₓ layers. SMIC uses a simpler local AlOₓ substack, while the remaining cap and etch sequence provide the required landing protection. So why double layers? The closely spaced AlOₓ doublets provide two protected endpoints in the etch sequence. Intel documents an AlOₓ/SiN/AlOₓ stack that explains the benefit. The main dielectric plasma etch stops on the first AlOₓ film; a selective wet clear opens that film; a second plasma etch removes the intermediate SiN and stops on the second AlOₓ film. The final wet clear exposes the metal landing surface. SiN is the intermediate dielectric in Intel’s published example. [8] The second stop protects the metal through a cap breakthrough. Wide openings can etch faster than narrow ones, and etch depth varies across the wafer. Metal under an early-clearing opening would otherwise be exposed while other openings still need more etching. Staged protection widens the process window and reduces metal erosion, corrosion and void formation. [8] TSMC documents AlN/AlOₓ/SiOC/AlOₓ above Cu, with AlN blocking Cu diffusion, and a simpler AlN/SiOC/AlOₓ variant that omits one AlOx film. [9] Levels with different opening sizes, aspect ratios, pattern densities and cap materials need different etch margins. A double AlOx stop is useful where another protected endpoint justifies the added processing. The extra film adds formation, selective opening and cleaning steps, plus another set of interfaces to control adhesion, moisture, and stress. These blanket films are opened through the existing via pattern, so each film does not require another lithography mask. AlOₓ adds parasitic capacitance when it replaces lower-k dielectric; two thin AlOₓ films can nevertheless contain less AlOₓ than one thick film. Total thickness, placement, and theintermediate dielectric determine the electrical cost. Deposition chemistry also changes AlOx permittivity and residual hydroxyl content, which can oxidize the underlying metal. [7], [8], [10], [11] The backside stack separates into relatively fine BM0-BM2 wiring near the devices and coarser BM3-BM5 power distribution. The largest pitch increase occurs between BM2 and BM3. BM0’s pitch closely matches the logic-row height, fitting local power delivery to the cell rows. Higher levels aggregate current through larger conductors: routing density becomes less important than low resistance and current capacity as the network approaches the package. This hierarchy provides wide power wiring for the power delivery network without consuming scarce frontside signal-routing resources. [3] The SemiAnalysis STEEL teardown lab breaks down advanced datacenter and AI hardware. To learn more about our pipeline or to commission a teardown, contact sales@semianalysis.com. WE’RE HIRING: Architecture, floorplan, packaging, manufacturing, and labs experts. Opportunities from system to transistor and everywhere in between. Check out our Careers page. Frontside Interconnects Intel 18A combines Mo-lined W contacts and nano-TSVs with a separate backside Cu power network. Samsung SF2 keeps power on the frontside, using Ti-based contact interfaces and Ta-based barriers and Co liners around Cu wiring. In 18A standard-cell rows, backside power rails supply the devices through nano-TSVs within the cells, freeing frontside routing resources. Samsung’s M0 accommodates both power and signal connections. From the device toward M0, the connection runs through a Ti-based S/D interface, W contact fill, a Mo-lined W via, and the Cu M0 wire. Mo supplies a conductive nucleation and adhesion layer for W, replacing the resistive TiN liner used in conventional W integration. This increases the effective conduction volume within the feature while retaining W fill and its established polishing, cleaning and etching processes. Intel’s Mo/W patent describes this integration tradeoff. The nano-TSV uses the same Mo-lined W construction in the backside supply path. [12] The move from TiN to Mo is an incremental change. While a full Co or Mo fill can also reduce the volume lost to liners in very small features, it requires new integration schemes that increase complexity and risk. Cu remains attractive for wider wires due to its low resistance. As wires and vias shrink, the diffusion barrier consumes an increasing fraction of their cross-section. [12], [12], [14] Intel uses Co/Ru liners at M0-M1, Co at M2-M4, and Nb at M5-M9. The lower-level liners help Cu adhere and reduce void formation during trench fills. Applied Materials’ Endura has new thermal control that facilitate wetting process, so the thin film continuity is good enough that good capillary pressure will drive Cu atoms to the via bottom without voiding. Intel’s choice to use Nb is particularly interesting. Intel’s Nb patent describes a conductive diffusion barrier intended to reduce the barrier’s contribution to resistance relative to conventional Ta-based barriers, particularly at via bottoms where all current crosses the barrier. The patent pairs Nb in coarser levels with the option of lower-cost PVD processing. [15], [16] The upper metal layers support thicker barriers formed through physical vapor deposition (PVD) despite its worse coverage and uniformity. Meanwhile, the lower metal layers require thinner barriers deposited through conformal atomic layer deposition (ALD). Co/Ru adds another material interface and requires controlled deposition and Cu fill. Changing liners and barriers by metal layer allows Intel to optimize interconnect resistance, process complexity, and reliability. [15, 16] RibbonFET, Intel’s name for its gate-all-around FETs (GAAFETs), replaces the FinFET’s vertical fins with four stacked horizontal silicon nanosheets, allowing the gate to surround the channel on every side. The path to GAAFET begins with the planar transistor. A planar MOSFET places the gate above the channel between its source and drain. Pairing an NMOS with a PMOS transistor creates a CMOS inverter, in which the NMOS pulls the output low for a high input, and the PMOS pulls it high for a low input. The gate must retain electrostatic control of the channel to ensure clean switching. As gate lengths shrank, the drain began to compete with the gate for that control, increasing off-state leakage. Electrostatic control was restored through an architectural evolution that raised the channel into a vertical fin and wrapping the gate around three sides. Called “FinFET”, this new architecture packed more effective channel width into a smaller footprint. Further scaling made it harder to maintain both drive current and leakage within smaller cells, and reintroduced the same problems planar MOSFETs faced. Nanosheet GAAFETs close the fourth side by replacing the vertical fin with a stack of horizontal nanosheets, each surrounded by the gate. The tighter electrostatic control suppresses leakage at shorter gate lengths while stacking adds effective channel width within the cell footprint. In a FinFET process, channel width changes in discrete steps as designers must add or remove whole fins. Nanosheet width can instead be adjusted continuously within the process’s design rules. Wider sheets increase drive current, while narrower sheets reduce capacitance at the cost of drive current. Intel 18A uses stacks of four nanosheets each and varies their widths across logic and SRAM. At the process level, adding more sheets to each stack increases effective channel width and drive current, but complicates fabrication. RibbonFET vs MBCFET Samsung began GAAFET production in 2022 with SF3E, following with SF3 and now SF2. Its ‘MBCFET’ provides a useful structural comparison with Intel’s first RibbonFET implementation. [17] STEEL is digging deeper into SF2, used in the Exynos 2600, and TSMC’s GAAFET N2, used in Apple’s A20 Pro, in upcoming newsletter articles. We’re throwing some teasers on X. Let’s compare Samsung SF2’s MBCFET with Intel 18A’s RibbonFET. Subscribe Even to the untrained eye, Intel’s extra nanosheet is obvious. Intel stacks four ribbons to Samsung’s three. Samsung’s sheets are much wider in these fields, so both sheet count and width matter to the available channel perimeter. Sheet width also changes which silicon surfaces carry current. On conventional (001) silicon, wide nanosheets emphasize the broad top and bottom surfaces, favoring electron transport; the larger sidewall contribution in a narrow sheet favors hole transport. Thinner sheets improve gate control but increase confinement and scattering. This makes width and thickness part of the NMOS/PMOS balance, alongside strain and threshold voltage. [18], [19] GAAFET designs like 18A use different work-function-metal (WFM) stacks for NMOS and PMOS. Around each ribbon, a thin SiOx interfacial layer separates the silicon channel from the HfOx high-k dielectric, with La providing dipole tuning and the WFM wrapping the dielectric. NMOS uses a TiAl-based stack, while PMOS uses TiN WFM. W fills the remaining gate trench, providing a lower-resistivity path where the work-function layers are no longer needed. In this field, the PMOS stacks leave room for W between ribbons, while the NMOS stacks occupy more of those gaps. A silicon-based dielectric marks the P/N boundary, allowing the PMOS and NMOS gates, sharing the same gate trench, to be processed sequentially. Fast logic paths, retention circuits, and SRAM need a family of threshold options. Changing threshold without substantially changing device dimensions, capacitance or fabrication complexity is valuable. FinFET processes typically use different work-function-metal stacks. In a four-ribbon GAA stack, the narrow sheet-to-sheet gap limits how much WFM can fit around each channel. La in the gate dielectric creates interfacial dipoles at the SiOx/HfOx boundary, shifting effective work function and tuning threshold voltage. This gives Intel another control alongside its NMOS and PMOS WFM stacks. Low-threshold devices improve critical-path drive; higher thresholds reduce leakage elsewhere. Dipole tuning is especially useful in GAA because it changes threshold without consuming the narrow intersheet gap with thicker WFM. Precise control of La incorporation, diffusion and interface quality has long been a challenge, limiting viability in high volume production but is now seen from every leading-edge foundry. Intel’s patent describes depositing a dipole-forming oxide above HfOx and annealing it toward the interfacial oxide before completing the work-function and fill metals. This separates threshold tuning from the space available for metal. Newer research addresses the thermal cost: imec’s 2026 dipole-middle research inserts the shifter between two HfOx depositions, shortening the diffusion path while protecting SiOx during patterning. [20], [21]Matched-cut EDS, Intel 18A (left) vs. Samsung SF2 (right). Intel retains raised source/drain epi beneath its contacts, while Samsung recesses W deep into the epi to form a V-shaped Ti-lined interface. The deeper contact increases metal-to-semiconductor area and shortens the current path from the lower sheets, reducing contact and spreading resistance. It also removes epi volume and brings the contact etch closer to the channel ends. Retaining more epi preserves the material available for strain transfer, especially from SiGe into PMOS. These geometries balance contact access against stress engineering and etch margin. [22], [23] Samsung stacks three sheets to Intel’s four ribbons, and both processes use sheet width to tune drive strength. In our Samsung cross-sections, widths range roughly from 19 to 30 nm in the NPU rows and 37 to 50 nm in the CU cell. The Samsung nanosheets taper, with the widest sheet at the bottom and the narrowest at the top. Both processes use HfOx gate dielectric and Ti-based work-function stacks, with Al in the NMOS stack. In the Samsung devices shown here, the dielectric and WFM occupy the intersheet gaps, leaving W above the top sheet. Intel’s PMOS stack leaves more room between ribbons, and W fills those gaps while the thicker NMOS stack leaves W mainly in the upper trench. Gate-stack EDS maps. The W between Intel’s PMOS ribbons provides a conductive path close to the lower gates. Where WFM fills the entire gap, the gate still surrounds the channel, but voltage reaches it through the more resistive work-function films. Thinner WFM and dipole tuning preserve room for low-resistivity fill; Mo and Ru are alternative fill metals being developed for further scaling. [24] A masked, sequential WFM flow explains the different gate heights and inter-nanosheet fill. The proposed sequence below shows how separate NMOS and PMOS work-function steps produce that geometry. Enabled by the BSPDN process, Intel replaces the dense-logic silicon subfin with dielectric, removing the parasitic conduction path below the ribbons and reducing substrate-related capacitance. A retained silicon body as in classical, non-SOI, planar and FinFET designs needs junction and punchthrough-stop engineering to suppress leakage. Dielectric isolation makes that leakage less sensitive to the subfin doping profile but adds removal and fill steps. It also weakens the direct thermal path through silicon, making the contacts, metal stacks and package more important for heat extraction. [24], [26] Fluorine is concentrated around selected Intel device structures in the maps. WF6 is a standard precursor for W nucleation and fill, while barrier films protect adjacent dielectrics from fluorine attack. Low-fluorine W processes reduce the residual-F burden. Chloride-based precursors avoid introducing F during W deposition, but require control of chlorine attack, nucleation and fill quality. The integration target is a continuous, low-resistance W path with a thin protective liner and minimal chemical damage to the surrounding stack. [20], [27], [28] The SemiAnalysis STEEL teardown lab breaks down advanced datacenter and AI hardware. To learn more about our pipeline or to commission a teardown, contact sales@semianalysis.com. WE’RE HIRING: Architecture, floorplan, packaging, manufacturing, and labs experts. Opportunities from system to transistor and everywhere in between. Check out our Careers page. We measured cell height, gate pitch, metal geometry, and ribbon dimensions at the XTEM sites shown below. The tables group these dimensions by site and device polarity. Our “sheet cuts” cross the silicon channel and show the ribbons end-on. “gate cuts” run along the channel through successive gates. The 18A logic cell dimensions point to a five-track logic library while the N3E and Intel 3 cell dimensions evidence a seven-track logic library. The DDR-PHY uses wider M0 wires and much larger spacing than core logic. That trades routing density for lower wire resistance and weaker coupling between neighboring nets. The geometry suits the current delivery and coupling requirements of analog, clock, and I/O circuitry. PowerVia lets 18A combine a compact cell height with wider M0 geometry by moving the main power rails off the signal-routing tracks. That relaxes local wire scaling while preserving a small cell footprint. Cell height and gate pitch set the geometric density; pin access and routability determine how much of it a real block can use. [29] The biggest takeaway from our gate-pitch measurements is that Intel 18A compute logic and TSMC N3E GPU logic have similar density in the Bohr representative-cell model. The 18A example is 18.6% denser than the Intel 3 GPU example. Gate pitches are nearly identical across the three sites, so cell height drives most of the difference. The Bohr model combines a four-transistor NAND2 spanning three gate pitches and a 32-transistor scan flip-flop (SFF) spanning nineteen pitches, weighting their densities 60:40. The sensitivity column shows how independently changing cell height and gate pitch by ±1 nm changes the result. This compares representative cell geometries; whole-die density also depends on cell mix and placement. The 18A P-core gives M0 substantially more metal cross section than the N3E vector engine. Treating each profile as a trapezoid gives 2.63 times the area per line and 1.84 times the area after normalization by routing pitch. The larger section reduces the geometric contribution to line resistance and lowers current density for a given current. Taller and wider wires also add capacitance, so circuit delay depends on the balance of resistance and capacitance. The DDR-PHY has less metal area per routing width than the 18A core fields, while remaining above N3E. [30] Area = height × (top CD + bottom CD) / 2, including liners. Area/pitch normalizes by routing width. Taper is the symmetric sidewall angle from vertical, with the largest angle belonging to the DDR-PHY. Compute tile The measurements show how ribbon dimensions and gate-stack geometry vary across the compute tile and between NMOS and PMOS to balance channel drive, gate load and the space needed for the dielectric/WFM stack across logic, SRAM and the DDR-PHY. Width mainly changes available channel perimeter; thickness also changes electrostatic control and carrier confinement. Gate-stack thickness then determines the space left for low-resistivity fill P-core and LP E-core logic Both the P-core and LP E-core use multiple nanosheet widths. Widths are measured on high-magnification XTEMs while wider-field images demonstrate additional width choices within the LP E-core. Multiple widths are expected even within an LP E-core. Timing-critical paths, buffers and cells with different fanout need different drive strengths. The lower-magnification fields show this width diversity beyond the sites quantified in the table. L2 and L3 SRAM GAA gives SRAM designers another way to balance the pull-up (PU), pass-gate (PG), and pull-down (PD) transistors. FinFET bitcells set device strength through fin count while GAA adds nanosheet width as a sizing knob. In a 6T SRAM cell, a strong pull-down relative to the pass-gate limits read disturbance, while a strong pass-gate relative to the pull-up improves writability. During a write, the pass-gate and write driver pull the node storing “1” below the inverter trip point. During a read, the pull-down holds the node storing “0” low. Bias, threshold voltage, mismatch and assist circuitry set the remaining margin. FinFET high-current cells commonly use a PU:PG:PD fin-count pattern of 1:2:2, a device-sizing ratio rather than a current ratio. Ribbon width lets Intel balance SRAM strengths without adding whole fins. The L2 cell uses its narrowest ribbons for PU and widest for PD, improving writability and read stability respectively. Intel’s disclosed HCC operates without assist; its denser HDC uses negative-bitline write assist. Pulling the selected bitline briefly below ground increases pass-gate overdrive so it can overpower the pull-up at lower supply voltage. That buys density and low voltage writability at the cost of boosting circuitry, switching energy, and additional voltage stress that must be controlled. [31], [32] Four rectangular ribbons give the perimeter = 8 × (width + thickness), before corner rounding. PG/PU is 1.49 and PD/PG is 1.16. The L3 structures closely resemble L2 in layout and cell height. Fewer L3 nanosheet widths are tabulated because fewer high-magnification images were available. DDR PHY The DDR-PHY trades density for controlled analog behavior and reliable off-chip signaling. It contains drivers, receivers, delay circuits, and calibration logic that set drive strength, sampling time, and voltage margin. Repeated four-sheet devices with similar widths fit the use of regular transistor units for matching and programmable drive. Its wider local wiring provides room for current delivery and separation of sensitive signals, while consuming more area than a dense core-logic grid. The layout serves the memory channel’s electrical requirements as well as digital logic density. [33] The SemiAnalysis STEEL teardown lab breaks down advanced datacenter and AI hardware. To learn more about our pipeline or to commission a teardown, contact sales@semianalysis.com. WE’RE HIRING: Architecture, floorplan, packaging, manufacturing, and labs experts. Opportunities from system to transistor and everywhere in between. Check out our Careers page. Intel 3 GPU devices Vector engine logic Intel 3’s XVE logic uses two-fin PMOS and NMOS devices with power rails in M0. Its cell height and M0 pitch give a seven-track geometry, two tracks more than the 18A logic. One-fin groups also appear among the two-fin devices. Intel 3 L2 SRAM The Intel 3 L2 SRAM uses the familiar HCC sizing pattern: one PU fin, two PG fins, and two PD fins. N3E GPU devices Vector engine logic The N3E XVE field contains repeated two-fin devices with seven-track cell geometry. N3E remains a FinFET process, giving Panther Lake a direct FinFET-to-RibbonFET comparison. N3E L2 SRAM The N3E L2 SRAM uses the same PU:PG:PD fin-count pattern of 1:2:2. Panther Lake-U follows Lunar Lake’s floorplan quite closely. Both pair 4 P-cores with 4 LP E-cores and NPU, media and display engines in similar locations. Lunar Lake also uses Xe2, the direct predecessor to Panther Lake’s Xe3 GPU. This makes Lunar Lake the most direct basis for our comparisons. Arrow Lake differs in core count and uses the older Xe-LPG GPU architecture, so we only use it where it offers a more direct component-level comparison. Compute tile Panther Lake compute-tile floorplans remain sparse even months after launch. Intel 18A’s backside metal and dielectric stack must be removed without damaging the underlying structures before a clean transistor-level floorplan can be imaged. Most published die shots hide or heavily process the background, but we are quite proud of the die shot we achieved and are excited to show the work we have done. We measured the areas of the key components on the compute tile and compared them with their Lunar Lake predecessors on TSMC N3B. These help us to capture changes in block area and compare the two chips across process nodes and designs. Our total tile areas exclude the scribe-line area. The compute-plus-GPU subtotal below uses the PTL-U compute tile and GT1 GPU; it excludes the I/O tile and passive base. Individual block areas use the boundaries marked on the floorplans The compute-plus-GPU row is recomputed from the displayed PTL-U and GT1 areas. Component rows use their stated per-region counts and are not an additive partition of the whole tile. The P-core area remains almost unchanged between Lunar Lake and Panther Lake, despite L2 capacity increasing from 2.5 MiB to 3 MiB. Arrow Lake uses the same Lion Cove core as Lunar Lake but also has a 3 MiB L2. Cougar Cove fits 20% more L2 into the same P-core area. The larger private cache keeps more of each core’s working set close to its execution units, reducing access to shared L3 and DRAM. Extra capacity adds storage leakage and lookup energy, so designers balance it against avoided lower-level accesses. The shared P-core L3 cache also shrank by 14.8%. [2] Cougar Cove combines a similar footprint with Intel’s reported power-efficiency improvements. RibbonFET’s tighter channel control reduces leakage, while PowerVia reduces supply droop and allows tighter voltage guardbands. [1] Darkmont’s four-core LP E-core cluster is 5.0% smaller than Skymont’s on Lunar Lake, with most of the reduction in its L2 regions. The 1 MiB region shrank by 8.4% and the 1.5 MiB region by 14.9%. The tag arrays also use one fewer visible row. Tags identify which memory addresses the data array holds, so rearranging them changes the cache’s layout and wiring without requiring less data capacity. [2] The LP E-cores share one L2. This pools capacity and avoids duplicating all the cache machinery, but the four cores contend for its banks and bandwidth. Their separate cluster also keeps light work away from the performance cluster and its L3, allowing that larger domain to sleep. [1], [2] Cache area includes more than the storage cells. Tags identify each line, decoders select rows, sense amplifiers read the small bitline signal, and wires connect to the banks. Splitting an array into smaller sections shortens wordlines and bitlines, improving access speed, but duplicates peripheral circuits. Panther Lake’s smaller cache regions therefore reflect the complete memory implementation, including how much of each region is devoted to storage. [34] Unlike Meteor Lake and Arrow Lake, Panther Lake has no separate SoC tile. The NPU, LP E-cores, memory controllers, PHYs, media and display engines now share the compute tile. This removes an active die and keeps CPU memory traffic on one die. The cost is moving PHY and I/O-related circuitry onto 18A: drivers, receivers and analog circuits must still meet external voltage, loading and signal-integrity requirements, so their area does not shrink like dense digital logic. [1], [2] The biggest shrink comes from the NPU, which occupies 36.9% less area. NPU 5 consolidates the same total INT8 MAC count into half as many neural compute engines. Each of the three NCEs has a larger MAC array to make the complete NCE envelope 22.6% larger than an NPU 4 engine. Consolidation also halves the number of scratchpads and SHAVE DSPs, from 12 to 6. The MAC array handles matrix multiplication and convolution, while SHAVE executes vector and custom operations that fit the array poorly. [1], [2], [35] The paired floorplans identify each NCE envelope and its scratchpad, MAC, and SHAVE regions. Each measured MAC polygon is counted once per NCE in the area accounting below. The scratchpads store weights, activations, and intermediate results near the MAC arrays, allowing repeated use without fetching them again from DRAM. Halving their number delivers the largest measured area saving but leaves less local storage for the same total MAC count. Layers that no longer fit locally require smaller working tiles or more transfers of intermediate data. The benefit depends on keeping the enlarged arrays busy while managing that tighter storage budget. [36] NPU 5 also adds native FP8. Using half the operand width of FP16 reduces storage and transfer demand, helping workloads fit the smaller local memory budget. Lower precision and format-dependent range make scaling and model validation part of deployment. Hardware activation functions further reduce work that would otherwise occupy the programmable DSPs. [1], [2] Microsoft requires an NPU to deliver at least 40 TOPS for Copilot+ PCs. Both Lunar Lake and Panther Lake meet this threshold, but Panther Lake uses significantly less silicon. GPU tiles Panther Lake is Intel’s first product with Xe3, its latest GPU architecture. It offers two different GPU tiles: a smaller GT1 tile with 4 Xe3 cores on Intel 3 and a larger GT2 tile with 12 Xe3 cores on TSMC N3E. Panther Lake allows us to compare the same GPU architecture across both Intel 3 and TSMC N3E. Wildcat Lake adds a third Xe3 implementation on Intel 18A. A future newsletter will detail Xe3 and its implementation differences across all three process nodes. GT2 scales Xe3 to a different physical layout, with render slices arranged vertically instead of GT1’s horizontal arrangement. Slice placement sets the distances to shared cache banks and the D2D interface. Those wires consume area and add delay, so scaling the number of Xe cores also requires a new balance of cache placement, routing and timing. [1] What’s immediately obvious is that the GT2 tile on TSMC N3E has much smaller Xe cores than GT1. These block areas include logic, caches, and routing. An Xe core on the GT1 tile is ~69% larger than one on Lunar Lake, and ~55% larger than one on GT2. Intel 3 therefore uses substantially more area per Xe core. The block-area gap exceeds the measured logic and SRAM density gaps, bringing routing, timing targets, cell mix, and floorplan allocation into the comparison. The measured vector/matrix engine region is almost unchanged between Lunar Lake and Panther Lake’s GT2 tile. Xe3 retains eight 512-bit vector engines and eight 2048-bit XMX engines per core. Its gains also come from feeding those engines more effectively: more resident threads hide stalls, and variable register allocation lets shaders trade registers per thread against the number of threads kept active. [1] The shared L1/SLM capacity increased by 33% from 192 KiB to 256 KiB, while its area increased only 5%, raising effective density by 27%. L1 retains reused cache lines, while software-managed SLM lets a thread group share data locally. Both reduce traffic to more distant memory. Allocating more SLM per group can also limit how many groups reside on a core at once. [1], [37] The GT1 tile carries 4 MiB of L2 against 16 MiB on the GT2 tile. GT1 divides its L2 cache into four 1 MiB banks, while GT2 uses eight 2 MiB banks. Each bank contains 128 macros, but each N3E macro stores 16 KiB, twice the Intel 3 macro’s 8 KiB capacity. The N3E macro is only 54% larger while holding twice as many bits, giving it 30% higher density: ~23.7 Mbit/mm² versus 18.3 Mbit/mm². Including bank-level circuitry, the gap widens to ~16.9 Mbit/mm² on GT2 versus ~10.4 Mbit/mm² on GT1. GT2 gains density with its macros storing more bits per unit area, and those macros occupy more of each cache bank. Larger macros spread decoder and sense-amplifier overhead across more storage, while a more compact bank layout reduces the share spent on control and routing. The compromise is longer wordlines and bitlines that carry more capacitance. [34] I/O tile Panther Lake uses two I/O tile variants, both fabricated on TSMC N6. The smaller one provides 4 PCIe 5.0 and 8 PCIe 4.0 lanes and serves lower-tier systems as well as those without a discrete GPU, while the larger one adds 8 PCIe 5.0 lanes, bringing the total to 20 lanes, for discrete-GPU connectivity. Panther Lake SKUs with the larger 10- or 12-Xe GPUs use the smaller I/O tile. [38] The smaller I/O tile adds a PCIe 4.0 block and a Thunderbolt block to Lunar Lake’s I/O layout, providing four additional PCIe 4.0 lanes and another Thunderbolt 4 port. Its repeated N6 blocks retain nearly identical areas and layouts. Reusing these proven PHYs and controllers avoids porting and requalifying external interfaces on 18A, where faster digital logic offers less benefit to circuits constrained by the off-chip link. [38] SemiAnalysis’s teardown lab (STEEL) dives deep into the world’s advanced datacenter and AI hardware. To learn more about our pipeline or to commission a teardown, contact sales@semianalysis.com. We’re hiring technical experts from system to transistor and everywhere in between. Check out our Careers page. Panther Lake offers scalability and modularity through its disaggregated packaging that partition compute, GPU, and I/O silicon into separate tiles allowing for a suite of tile configurations. This partitioning makes the package part of Intel’s node economics as it determines how much leading-edge wafer area each product consumes, which functions can remain on other processes, and how much configuration freedom Intel can offer from a shared set of tiles. Furthermore, fabricating the compute and GPU tiles separately confines the new 18A process to the compute tile and allows graphics and I/O to use other, more established, and more cost-effective processes. For Panther Lake, the GPU and I/O tiles are assembled alongside the compute tile on a passive silicon base using Foveros-S. Intel’s current technology brief lists a nominal 36 µm pitch for Foveros-S. Through-silicon vias (TSVs) in the base connect the fine wiring above to the larger package connections below. The functional tiles sit side by side on that passive base in a 2.5D configuration. [39] Our cross-section through the compute and GPU tiles shows the package’s wiring hierarchy. Microbumps connect each active tile to the passive silicon base; its fine redistribution layer (RDL) carries the short, dense tile-to-tile links. TSVs carry connections through the base to the package substrate, which fans them out to the much coarser motherboard solder joints. The base supplies interconnect, while computation remains in the active tiles above it. [39] At the compute-tile edge, the higher-magnification inset shows a local microbump spacing of approximately 25.24 µm and a feature width of 12.33 µm. These local spacings are finer than Intel’s nominal Foveros-S value. The X-ray fields further confirm tighter neighboring bumps, consistent across every die-to-die area found on each tile. Additional X-ray analysis is offered after the paywall. Putting the memory controller beside the CPU removes the D2D transfer that CPU memory requests required in Meteor Lake and Arrow Lake. This avoids the extra transmitter, receiver, and link traversal, saving interface energy and latency. Panther Lake’s separate GPU still crosses a D2D link to reach DRAM, so its larger local caches also help contain package traffic. [1], [40] Smaller dies are less likely to contain a random fatal defect, and screening them before assembly prevents one bad tile from consuming a complete package of good silicon. Reuse also spreads design and qualification work across more products. Against those gains, Intel pays for the passive base, D2D circuits, extra bonding and test steps, and losses during assembly. Cost per working product across the portfolio captures the combined effect of wafer yield, reuse, test, and assembly. [29] Wildcat Lake packaging Intel launched Core Series 3, formerly Wildcat Lake, on 16 April 2026 for value mobile and edge systems. Wildcat Lake keeps 18A but removes the passive base and combines more functions on one die to simplify the package. The two products therefore reveal two distinct ways to commercialize the same leading-edge process. [41] Wildcat Lake’s 18A die combines up to two Cougar Cove P-cores, four Darkmont LP E-cores, two Xe3 cores and a smaller NPU. A separate platform-controller die supplies I/O, connected through UCIe, Intel’s first processor implementation of the standard. Consolidating graphics remove a tile boundary and the passive base, reducing assembly complexity for a modest-bandwidth value product. It also ties CPU and graphics scaling to the same die, giving up Panther Lake’s ability to swap in a much larger GPU. [42], [43] In July 2021, Intel CEO Pat Gelsinger set out an ambitious process roadmap aimed at regaining performance leadership by 2025, later described as five nodes in four years. Five years and one CEO later, Intel’s comeback story is not as unambiguously positive as Pat may have hoped. [44], [45] Intel once set the pace for process technology, bringing high-k metal gate technology and FinFETs into volume production years ahead of the rest of the industry. Its 22 nm FinFET process reached consumers with Ivy Bridge in 2012. [46] Intel’s integrated device manufacturing (IDM) model allowed its architects and process engineers to co-optimize products and processes. Starting with Sandy Bridge, Intel dominated x86, while AMD struggled with Bulldozer. That lead faltered at 14 nm and broke at 10 nm. Intel targeted a massive 2.7× density increase, but the node arrived years late and required several revisions before it could support Intel’s full lineup. This delay forced Intel to stretch 14 nm across six generations, while TSMC moved ahead in process technology and AMD recovered in x86. By 2019, Intel was still shipping 14 nm across most of its product stack, with its 10 nm client ramp focused on Ice Lake mobile processors. Meanwhile, TSMC was shipping N7 and N7+, and AMD’s Zen 2 compute chiplets used N7 to raise core counts and improve efficiency. Intel’s process failures were central to its decline, but unsound business decisions furthered their downward slide. Product delays compounded product mistakes, pushing client, server, and FPGA roadmaps off schedule. Several attempts to enter AI (Nervana and Gaudi) and networking (Tofino) also failed to establish lasting businesses. Intel’s recovery has focused on consumer CPUs and advanced packaging. Tiger Lake, Alder Lake, Lunar Lake and now Panther Lake have restored Intel’s consumer roadmap. On the process side, Intel 4 shipped with Meteor Lake, Intel 3 with Granite Rapids and Sierra Forest, and Intel 18A with Panther Lake. Intel has also made advanced packaging part of its foundry offering. However, Intel is still playing catch-up in servers. Several Xeon generations arrived years late and trailed contemporary AMD and Arm server CPUs in performance, efficiency, and core count. The process roadmap is back, but Intel does not hold the same process-technology leadership position it held prior to 10 nm. The introduction of gate-all-around nanosheets and backside power delivery are two of the biggest changes to transistor integration in a decade. Intel took on both changes at once: 18A paired its first RibbonFET with PowerVia in Panther Lake. Panther Lake is a substantial manufacturing milestone. Our cross-sections show how RibbonFET and PowerVia reshape local contacts and wiring, while the floorplans show where architectural consolidation and process choices save area. A sustained competitive lead depends on product performance, cost, yield, and the next implementation. The SemiAnalysis STEEL teardown lab breaks down advanced datacenter and AI hardware. To learn more about our pipeline or to commission a teardown, contact sales@semianalysis.com. WE’RE HIRING: Architecture, floorplan, packaging, manufacturing, and labs experts. Opportunities from system to transistor and everywhere in between. Check out our Careers page.. [1] Intel, “Intel Tech Tour 2025 Panther Lake recap,” Doc. 866361, 2025. Accessed: Sep. 14, 2026. [Online]. Available: https://cdrdv2-public.intel.com/866361/ITT_2025_Panther_Lake_Recap1.pdf [2] Intel, “Core Ultra processors Series 3 datasheet, volume 1,” Doc. 872188-001, Jan. 2026. Accessed: Sep. 14, 2026. [Online]. Available: https://cdrdv2-public.intel.com/872188/872188-001.pdf [3] N. Horiguchi and E. Beyne, “Backside power delivery: How to power chips from the backside,” imec Reading Room, Nov. 25, 2022. Accessed: Sep. 14, 2026. [Online]. Available: https://www.imec-int.com/en/articles/how-power-chips-backside [4] W. Hafez et al., “Intel PowerVia technology: Backside power delivery for high density and high-performance computing,” in Proc. IEEE Symp. VLSI Technol. Circuits, Kyoto, Japan, Jun. 2023, pp. 1–2, doi: 10.23919/vlsitechnologyandcir57934.2023.10185208. [5] K. Fischer et al., “Intel 18A platform technology featuring RibbonFET (GAA) and PowerVia for advanced high-performance computing,” in Proc. Symp. VLSI Technol. Circuits, Kyoto, Japan, Jun. 2025, pp. 1–3, doi: 10.23919/vlsitechnologyandcir65189.2025.11075006. [6] K.-F. Cheng, C.-L. Teng, H.-Y. Huang, and H.-C. Chen, “Metal oxide composite as etch stop layer,” U.S. Patent 12 176 247 B2, Dec. 24, 2024. Accessed: Sep. 14, 2026. [Online]. Available: https://patents.google.com/patent/US12176247B2/en [7] S. W. King, “Dielectric barrier, etch stop, and metal capping materials for state of the art and beyond metal interconnects,” ECS J. Solid State Sci. Technol., vol. 4, no. 1, pp. N3029–N3047, 2015, doi: 10.1149/2.0051501jss. [8] A. V. Mule’, D. J. Towner, D. Seghete, C. R. Ryder, and A. A. Gonzalez, “Multi-layer etch stop layers for advanced integrated circuit structure fabrication,” U.S. Patent Appl. 20220102343 A1, Mar. 31, 2022. Accessed: Sep. 14, 2026. [Online]. Available: https://patents.google.com/patent/US20220102343A1/en [9] C.-C. Wang and J. H. Wang, “Interconnect structure with dielectric cap layer and etch stop layer stack,” U.S. Patent Appl. 20230253247 A1, Aug. 10, 2023. Accessed: Sep. 14, 2026. [Online]. Available: https://patents.google.com/patent/US20230253247A1/en [10] Y. J. Wu, K.-F. Cheng, C.-C. Lee, H.-K. Chang, and H.-Y. Huang, “Etch stop layer for interconnect structures,” U.S. Patent Appl. 20240332070 A1, Oct. 3, 2024. Accessed: Sep. 14, 2026. [Online]. Available: https://patents.google.com/patent/US20240332070A1/en [11] M. G. Rainville, N. Shankar, K. S. Reddy, and D. M. Hausmann, “Deposition of aluminum oxide etch stop layers,” U.S. Patent Appl. 20180197770 A1, Jul. 12, 2018. Accessed: Sep. 14, 2026. [Online]. Available: https://patents.justia.com/patent/20180197770 [12] J. S. Leib et al., “Conductive lines having molybdenum liner and tungsten fill for advanced integrated circuit structure fabrication,” U.S. Patent Appl. 20240429126 A1, Dec. 26, 2024. Accessed: Sep. 14, 2026. [Online]. Available: https://patents.justia.com/patent/20240429126 [13] M. Naik, “Cobalt enables power and performance scaling at single-digit logic nodes,” Applied Materials, Dec. 17, 2018. Accessed: Sep. 14, 2026. [Online]. Available: https://www.appliedmaterials.com/us/en/blog/blog-posts/cobalt-enables-power-and-performance-scaling-at-single-digit-logic-nodes.html [14] Lam Research, “Breaking through AI’s invisible barrier with molybdenum,” Lam Research Newsroom, May 6, 2025. Accessed: Sep. 14, 2026. [Online]. Available: https://newsroom.lamresearch.com/molybdenum-metallization-ai-revolution [15] P. Yashar, G. Malyavanatham, and H. Vijwani, “Integrated circuit interconnect structures with niobium barrier materials,” U.S. Patent Appl. 20240112951 A1, Apr. 4, 2024. Accessed: Sep. 14, 2026. [Online]. Available: https://patents.google.com/patent/US20240112951A1/en [16] Applied Materials, “Applied Materials unveils chip wiring innovations for more energy-efficient computing,” Applied Materials Investor Relations, Jul. 8, 2024. Accessed: Sep. 14, 2026. [Online]. Available: https://ir.appliedmaterials.com/news-releases/news-release-details/applied-materials-unveils-chip-wiring-innovations-more-energy [17] Samsung, “Samsung showcases AI-era vision and latest foundry technologies at SFF 2024,” Samsung Semiconductor, Jun. 13, 2024. Accessed: Sep. 14, 2026. [Online]. Available: https://semiconductor.samsung.com/news-events/news/samsung-showcases-ai-era-vision-and-latest-foundry-technologies-at-sff-2024/ [18] C. W. Yeung et al., “Channel geometry impact and narrow sheet effect of stacked nanosheet,” in Proc. IEEE Int. Electron Devices Meeting (IEDM), Dec. 2018, pp. 28.6.1–28.6.4, doi: 10.1109/iedm.2018.8614608. [19] S. Mochizuki et al., “Evaluation of (110) versus (001) channel orientation for improved nFET/pFET device performance trade-off in gate-all-around nanosheet technology,” in Proc. Int. Electron Devices Meeting (IEDM), Dec. 2023, pp. 1–4, doi: 10.1109/iedm45741.2023.10413854. [20] D. G. Ouellette et al., “Fabrication of gate-all-around integrated circuit structures having molybdenum nitride metal gates and gate dielectrics with a dipole layer,” U.S. Patent 12 051 698 B2, Jul. 30, 2024. Accessed: Sep. 14, 2026. [Online]. Available: https://patents.google.com/patent/US12051698B2/en [21] imec, “Advancing the CFET-based device roadmap with novel integration modules and standard-cell architectures,” imec, 2026. Accessed: Sep. 14, 2026. [Online]. Available: https://www.imec-int.com/en/articles/advancing-cfet-based-device-roadmap-novel-integration-modules-and-standard-cell [22] A. Reznicek, X. Miao, C. Lee, and J. Zhang, “Wrap around contact for nanosheet source drain epitaxy,” U.S. Patent 11 302 813 B2, Apr. 12, 2022. Accessed: Sep. 14, 2026. [Online]. Available: https://patents.google.com/patent/US11302813B2/en [23] J. A. Kittl, J. G. Hong, D. R. Palle, and M. S. Rodder, “Methods for varied strain on nano-scale field effect transistor devices,” U.S. Patent 9 978 833 B2, May 22, 2018. Accessed: Sep. 14, 2026. [Online]. Available: https://patents.google.com/patent/US9978833B2/en [24] S. Gandikota et al., “Method of reducing metal gate resistance for next generation NMOS device application,” Int. Patent Appl. WO 2024/137272 A1, Jun. 27, 2024. Accessed: Sep. 14, 2026. [Online]. Available: https://patents.google.com/patent/WO2024137272A1/en [25] J. Zhang et al., “Full bottom dielectric isolation to enable stacked nanosheet transistor for low power and high performance applications,” in Proc. IEEE Int. Electron Devices Meeting (IEDM), Dec. 2019, pp. 11.6.1–11.6.4, doi: 10.1109/iedm19573.2019.8993490. [26] C. Yoo, J. Chang, Y. Seon, H. Kim, and J. Jeon, “Analysis of self-heating effects in multi-nanosheet FET considering bottom isolation and package options,” IEEE Trans. Electron Devices, vol. 69, no. 3, pp. 1524–1531, Mar. 2022, doi: 10.1109/ted.2022.3141327. [27] S. S. Pradhan, D. B. Bergstrom, J.-S. Chun, and J. Chiu, “Tungsten gates for non-planar transistors,” European Patent Appl. EP 3 506 367 A1, Jul. 3, 2019. Accessed: Sep. 14, 2026. [Online]. Available: https://patents.google.com/patent/EP3506367A1/en [28] L. Schloss and X. Ba, “Tungsten films having low fluorine content,” U.S. Patent 9 754 824 B2, Sep. 5, 2017. Accessed: Sep. 14, 2026. [Online]. Available: https://patents.google.com/patent/US9754824B2/en [29] Intel Foundry, “Accelerating AI and HPC with advanced process and packaging technologies,” Doc. 367370-001US, Jun. 2026. Accessed: Sep. 14, 2026. [Online]. Available: https://www.intel.com/content/dam/www/central-libraries/us/en/documents/2025-11/intel-foundry-hpc-ai-brief.pdf [30] M. Lofrano, X. Chang, H. Oprins, and Z. Tokei, “Mitigating the thermal bottleneck in advanced interconnects,” imec, Sep. 28, 2023. Accessed: Sep. 14, 2026. [Online]. Available: https://www.imec-int.com/en/articles/mitigating-thermal-bottleneck-advanced-interconnects [31] S. Bair, “Intel at ISSCC 2025: Navid Shahriari invited talk, eight papers, forums, panelist & product details,” Intel Community, Feb. 19, 2025. Accessed: Sep. 14, 2026. [Online]. Available: https://community.intel.com/t5/Blogs/Tech-Innovation/Edge-5G/Intel-at-ISSCC-2025-Navid-Shahriari-Invited-Talk-Eight-Papers/post/1667592 [32] D. Chandra, E. Potladhurthi, D. R. S. Reddy, and K. S. Rengarajan, “Tunable negative bitline write assist and boost attenuation circuit,” U.S. Patent Appl. 20160203857 A1, Jul. 14, 2016. Accessed: Sep. 14, 2026. [Online]. Available: https://patents.google.com/patent/US20160203857A1/en [33] Synopsys, “Advantages of firmware-based training in high-speed DDR IP,” Synopsys IP Technical Bulletin. Accessed: Sep. 14, 2026. [Online]. Available: https://www.synopsys.com/articles/firmware-based-training-ddr-ip.html [34] N. Muralimanohar, R. Balasubramonian, and N. P. Jouppi, “CACTI 6.0: A tool to understand large caches,” Hewlett-Packard Laboratories, 2009. Accessed: Sep. 14, 2026. [Online]. Available: https://users.cs.utah.edu/~rajeev/cacti6/cacti6-tr.pdf [35] Intel, “Intel Tech Tour 2024 Lunar Lake AI hardware accelerators,” Doc. 824436, 2024. Accessed: Sep. 14, 2026. [Online]. Available: https://cdrdv2-public.intel.com/824436/2024_Intel_Tech%20Tour%20TW_Lunar%20Lake%20AI%20Hardware%20Accelerators.pdf [36] Y.-H. Chen, J. Emer, and V. Sze, “Eyeriss: A spatial architecture for energy-efficient dataflow for convolutional neural networks,” in Proc. 43rd ACM/IEEE Annu. Int. Symp. Comput. Archit. (ISCA), Jun. 2016, pp. 367–379, doi: 10.1109/isca.2016.40. [37] Intel, “Introduction to the Xe-HPG architecture,” Doc. 758306, Nov. 4, 2022. Accessed: Sep. 14, 2026. [Online]. Available: https://www.intel.com/content/www/us/en/developer/articles/technical/introduction-to-the-xe-hpg-architecture.html [38] Intel, “Core Ultra processors Series 3 for the edge,” Doc. 855291. Accessed: Sep. 14, 2026. [Online]. Available: https://cdrdv2-public.intel.com/855291/Intel%C2%AE%20Core%E2%84%A2%20Ultra%20Processors%20Series%203%20for%20Edge%20Overview_2.pdf [39] Intel Foundry, “Foveros technology brief,” Doc. 366411-001US, Feb. 2026. Accessed: Sep. 14, 2026. [Online]. Available: https://www.intel.com/content/dam/www/central-libraries/us/en/documents/2025-07/foveros-25d-product-brief.pdf [40] D. Das Sharma, “UCIe: Building an open chiplet ecosystem,” UCIe Consortium, 2022. Accessed: Sep. 14, 2026. [Online]. Available: https://www.uciexpress.org/_files/ugd/0c1418_c5970a68ab214ffc97fab16d11581449.pdf [41] Intel, “Intel launches Intel Core Series 3 processors,” Intel Newsroom, Apr. 16, 2026. Accessed: Sep. 14, 2026. [Online]. Available: https://www.intel.com/content/www/us/en/newsroom/news/client-computing/intel-launches-intel-core-series-3-processors-changing-the-game-for-everyday-computing.html [42] Intel, “Core Series 3 launch press deck,” Apr. 2026. Accessed: Sep. 14, 2026. [Online]. Available: https://download.intel.com/newsroom/2026/Intel-Core-Series-3/Intel-Core-Series-3-Launch-Press-Deck.pdf [43] Intel, “Intel outlines architectures for agentic AI at Hot Chips 2026,” Intel Newsroom, Aug. 24, 2026. Accessed: Sep. 14, 2026. [Online]. Available: https://www.intel.com/content/www/us/en/newsroom/news/client-computing/intel-outlines-architectures-for-agentic-ai-at-hot-chips-2026.html [44] Intel, “Intel accelerates process and packaging innovations,” Intel Investor Relations, Jul. 26, 2021. Accessed: Sep. 14, 2026. [Online]. Available: https://www.intc.com/filings-reports/all-sec-filings/content/0001193125-21-224438/d199788dex991.htm [45] Intel, “Intel reports third-quarter 2021 financial results,” Intel Investor Relations, Oct. 21, 2021. Accessed: Sep. 14, 2026. [Online]. Available: https://www.intc.com/news-events/press-releases/detail/1505/intel-reports-third-quarter-2021-financial-results [46] Intel, “3rd generation Intel Core processors bring exciting new experiences and fun to the PC,” Intel Investor Relations, Apr. 23, 2012. Accessed: Sep. 14, 2026. [Online]. Available: https://www.intc.com/news-events/press-releases/detail/577/3rd-generation-intel-core-processors-bring-exciting
Remedy has confirmed performance issues with AMD graphics cards in CONTROL Resonant.
Developers from Remedy Studio have officially confirmed technical issues with CONTROL Resonant for AMD graphics card users. Players...
Wici Wireless eGPU Promises RTX 5060 Ti and 5090 Performance Over Wi-Fi 7
There has been a lot of development in the eGPU scene lately, with Thunderbolt 5, OCuLink, and MCIO 8i providing serious gains in bandwidth and throughput for eGPU use cases. However, Wici has just launched the Wici One, a new wireless eGPU solution that relies on Wi-Fi 7 for data transfer...
Google's $5 AI plan is now even more tempting
Google AI plans now include premium Colab benefits, offering faster computing, GPU access, and more for subscribers.
First wireless eGPU with RTX 5060 Ti set to ship this year, Cyberpunk 2077 demo shown
WiCi One wants to make the eGPU cable optional. The RTX 5060 Ti box uses Wi-Fi 7 and a virtual GPU driver, with shipping planned this year.
Early Reality Fracture Data Highlights Biggest Traps To Avoid
Thanks to our own internal data, we've been able to put together a picture of Reality Fracture, and it seems like some cards are being way over-valued by the community. Are you taking any of these trap cards?
CyberPowerPC Gaming PCs With Founder's Edition Nvidia GeForce RTX 5070 GPUs Are on Sale During the Walmart GeForce Week Sale Event
Part of the Walmart's GeForce Week Sale that ends on September 27.
Quanome Advances AI Cloud Business with $18.8 Million GPU Infrastructure Purchase Agreement
Quanome Technologies (NASDAQ:QNME) has entered into an approximately $18. 8 million GPU infrastructure purchase agreement for its XDT AI cloud business, with its first U.
Quanome Advances Its AI Cloud Business XDT with US$18.8 Million Infrastructure Purchase Agreement
Company Advances Development of First U.S. AI Infrastructure Hub, Executes Its Quantum & AI Systems Strategy ITASCA, Ill., Sept. 23, 2026 (GLOBE NEWSWIRE) -- Quanome Technologies, Inc. (Nasdaq: QNME) (“Quanome” or the “Company”) is advancing its Quantum and AI Systems business through an approximately US$18.8 million purchase agreement for computing infrastructure powered by NVIDIA GPU technology to build its AI cloud business through its subsidiary, XDT Infrastructure I, LLC (“XDT”). The Compan
PC version ControlResonant leaked a day before its official release.
A computer version of a sci-fi action film Control: Resonant from Finnish studio Remedy Entertainment has become freely available on...
CyberPowerPC Launches RTX Founders Edition Pre-Built PCs Starting at $1,549
NVIDIA's Founders Edition graphics cards are generally restricted to its own retail channels, but it seems as though CyberPowerPC has a limited run of RTX 50-series GPUs available in its pre-built gaming PCs for a GeForce Week event, which is slated to run until September 27. The CyberPowerPCs...
CyberPowerPC now sells prebuilt PCs with NVIDIA RTX 50 Founders Edition GPUs, 1,000 RTX 5080 systems sell in one day
CyberPowerPC is using NVIDIA's own RTX 50 Founders Edition cards in new Walmart systems. The RTX 5080 configuration with Ryzen 7 9800X3D costs $2,299.
GPUS Stock Slips After Hyperscale Data Carves Out Bitcoin Mining Business Into New Unit
Hyperscale Data management said Bitcoin mining and AI infrastructure require different capital strategies and priorities.
Micron reportedly ends production of 2GB GDDR7 used by GeForce RTX 50
Micron appears to be moving away from 2GB GDDR7 as its catalog shifts toward 3GB devices. The older 16Gb parts were already used on RTX 50 cards.
Micron 2 GB GDDR7 Memory Chips Reach End Of Life, Leaving RTX 50 Series Fully Dependent On Samsung And SK Hynix
Micron halts production of 2 GB GDDR7 memory chips, impacting RTX 50 GPU availability; 3 GB modules deemed largely ineffective.
Original ‘Lioness’ star opens up about her shocking return in the season 3 finale: ‘I screamed’
The ‘Lioness’ actor behind the surprise cameo at the end of the season 3 finale is sharing the details of her shocking return.
Walmart Has the Best Deal on a GeForce RTX 5080 Prebuilt Gaming PC I've Seen in Months, and It Even Includes a Founder's Edition GPU
Equipped with an RTX 5080 Founder's Edition GPU, 32GB of RAM, and a 1TB SSD.
Which Nvidia GPU Is Equivalent To The PlayStation 5 Pro?
While the PlayStation 5 Pro is the most graphically capable console, its PC equivalent GPU has some other tricks up its sleeve that make it just as capable.
M5 Ultra and M6 Chip Benchmark Results Reveal Graphics Performance
The first GPU benchmark results for Apple's new M5 Ultra and M6 chips have surfaced, ahead of the new Mac Studio and Mac mini models launching on Tuesday. If you are interested in CPU results, we shared those earlier this week. So far, the M5 Ultra has achieved up to a 59% higher Metal score (366,744) on Geekbench 7 compared to the M3 Ultra's average score of 230,420, and the chip's OpenCL score is roughly in line with NVIDIA's GeForce RTX 4080.
World of Warcraft: Forever drops GeForce GTX 780 support, but 2012 Radeon HD 7850 is still supported
World of Warcraft: Forever raises its official GPU minimum to GTX 10 series and RDNA 1. A separate architecture cutoff still allows some much older GPUs.
Engrams Embedding Entendre: Codesign for Efficient DRAM/SSD Offloading
Engram extends standard token embeddings with learned multi-token lookups. Recurring local patterns retrieve vectors directly, reducing the need to reconstruct them through attention and feed-forward layers. With Engram model architecture optimization, it allows for lower HBM capacity to be needed for models at the same quality. This does not mean there won’t be an insane demand for HBM but it just means that model architecture will continue to innovate around constraints. This model architecture design is naturally codesigned for parameter offloading: each token accesses a few embedding rows whose addresses depend on token IDs, not hidden states. The runtime can prefetch those rows from host DRAM while earlier layers compute, keeping the table outside HBM without transferring entire weight matrices. Our Memory model contains our latest estimates of quarter by quarter HBM, DRAM, & NAND supply and demand. Offloading frees HBM for model weights and KV cache, potentially supporting larger batches or more concurrent sessions. When DRAM becomes the next constraint, NVMe offers another tier. Recommendation systems already cache frequently or recently accessed embedding rows in faster memory while backing colder rows with SSDs. After NVIDIA roadmap had to change due to massively despec’ing Rubin Ultra from 1024GB to now ~200GB of HBM per chip, model architecture optimizations like emgram maybe helpful. Our DeepSeek-V4.1-Flash configuration uses roughly 189 GiB of memory for Engram. We replace it with a memory-mapped (mmap) file and measure serving performance with offload to SSD. Later on in our report, we will show our Engram offloading experiments along with the official InferenceX agentic inference serving results on Engram models like DeepSeekv4.1 Flash across all 6 NVIDIA GPU SKUs along with MI355X. Unsurprisingly, the CUDA Moat is still mogging MI355X on the ultra popular DeepSeekV4.1 Flash model. We also show how even on high capacity HBM SKUs, offloading emgrams to DRAM could result in even better performance for most of the pareto than keeping the emgram in HBM. Our benchmark has been widely reproduced, validated and/or supported by almost every major buyer of compute from Google Cloud to Microsoft Azure to Oracle, to Meta and many more. Furthermore, it has the support of the ML community including from vLLM, LMCache, SGLang, PyTorch, Huggingface and the support of major labs like OpenAI, MiniMax, ZAI, Qwen, Moonshot Kimi, etc. Star the InferenceX GitHub repository if you find the open-source benchmark and data useful!. InferenceX is the only inference benchmark in the world to have TPUv7, Jalapeño, Nvidia Rubin NVL72, AMD, and soon, SambaNova and Trainium. Due to how realistic AgentX scenario is to real world agentic inference workloads, AMD has committed to collaborating on MI455X UALoE72 too. DeepSeek did not release the original paper’s two trained Engram models. We replicated its setup on fineweb-edu using the released code and training hyperparameters, at an estimated 6E18 FLOPs per run. We observed the same U-shape scaling: Engram improved performance over pure MoE baselines. We also reproduced DeepSeek’s results, where earlier-layer representations with Engram resembled those of later layers. Like the original Engram paper, it is possible to probe Engram’s gate scores to see what n-grams DeepSeek-V4.1-Flash makes the most use of. Our gate scan finds names, code fragments, relational phrasing, and boilerplate. These examples prioritize interesting-ness over gate strength. One unexpected result was Wright : Ace Attorney. These examples suggest learned memory optimizes the training objective, not a judgment of which facts deserve storage. Licenses, bibliography fragments, API scaffolding, and website furniture can provide prediction shortcuts, so the value of additional Engram capacity may depend on what survives data preparation. This does not show that table capacity is “wasted”: the evaluation-corpus scan establishes neither training exposure nor the capacity occupied by each category. For offloading, strong gates do not identify cache-hot rows. Low gates do not automatically save reads either: computing the gate requires the retrieved key and undoes the performance gain of a fused kernel. Skipping reads would require a separate usefulness predictor before retrieval. In the original paper’s inference-time ablation, factual-knowledge benchmarks retained just 29–44% of their original performance, while reading comprehension retained 81–93%. This is due to the training–inference mismatch. The resulting degradation therefore measures this trained model’s dependence on Engram, not the performance difference between models trained with and without it. In our ablations, suppressing Engram worsens token likelihood across all evaluated domains, especially encyclopedia text and several code corpora. Surprisingly, GSM8K accuracy stays within measured run-to-run variation and removing Engram has no effect. Engram is not a detachable dictionary beside an unchanged MoE. Removing it changes downstream features and expert selection. We tested whether rerouting hurts or compensates by holding tokens fixed in a teacher-forced experiment on CRUXEval, a code-reasoning benchmark of small Python functions where the model predicts a function’s output from its code and an input, and scoring the reference answer. Removing Engram raised answer loss from 0.2848 to 0.3093 bits/token. Forcing the ablated model to use the original Engram-on expert choices made it worse still, at 0.3375 bits/token. Rerouting partially compensates for the missing memory. Memory features and expert selection work together, rather than following a clean “memory stores facts; experts reason” division. On the same CRUXeval, removing Engram during either phase reduced accuracy and increased generated tokens; removing it throughout produced the largest changes. Keeping Engram for prefill leads to more correct answers than keeping for decode likely due to semantically richer KV cache transferred to decode workers, allowing it to mitigate some of the performance loss. Engram’s table is large, but each lookup is small. DeepSeek-V4.1-Flash requests 24 rows at each of two Engram layers, about 12.4 KiB per processed token position across the model, or 3.1 KiB per GPU when split across four GPUs. Currently as of Day 7 since Model Release, MI355X is still 2-4x worse performance per dollar compared to B200 even when normalized by Mi355X’s lower TCO. Our full total cost of ownership breakdown comes from our AI Cloud TCO Model along with monthly market surveys of over 100+ gpu clouds & gpu cloud customers. On the Day 0 release of DeepSeekv4.1 Flash, NVIDIA vLLM works out of the box with zero issues across all 6 SKUs: H100, H200, B200, B300, GB200, GB300! This was thanks to the amazing work by the NVIDIA & Interact teams! In comparison, AMD vLLM did not work on day 0 for DeepSeekv4.1 Flash.Source: SemiAnalysis InferenceX AMD’s vLLM documentation points to using vllm/vllm-openai-rocm:deepseekv41-flash-0909, but from hour 0 of the model release to hour 23, AMD has not publicly released the image. AMD claims, “SPEED IS THE MOAT,” yet it has still not released it by the 23rd hour. We wish that, going forward, the AMD team has a better process for hour 0 model releases. Eventually when they did publicly release for “day 0” image support, performance-wise, it is currently up to 14.8x worse perf per dollar than H200 and up to 42x worse perf per dollar than B200/B300. The power of the CUDA MOAT is NVIDIA’s collaboration with its massive 6 million-developer community ecosystem including most of the vLLM & SGLang & Tokenspeed maintainers which means that CUDA is optimized on day 0. Overall, AMD did make significant improvements but the performance per dollar is still currently 2-4x worse than B200. AgentX Engram DRAM Offloading Improving Performance The HBM and DRAM offload use the same GPU kernel to select and dequantize rows. With HBM, it reads GPU memory; with Unified Virtual Addressing (UVA), it reads pinned host memory directly. Both support full decode graphs. Moving the table into HBM accelerates only the sparse lookup, leaving decoder computation and communication unchanged, resulting in little overall benefit while consuming memory otherwise available to KV cache. Another benefit of Engram offloaded to DRAM which means you can reduce the communication overhead by using less HBM GPUs per replica. For example, when enabling Engram offloading on B300, we are able to switch from TP4 to now TP2 which improves the pareto curve by up to 1.6x. When iso-model quality, less HBM is required as the engrams could be offloaded to host DRAM. Thus HBM bandwidth matters way more than HBM capacity. For inference workloads where memory bandwidth matters the most, 4-hi HBM provides the best $/bandwidth and therefore lowest cost per token. If China continues to make more and more revolutionary model architecture innovations, soon it could potentially 0Hi HBM stacks. [ ](https://newsletter.semianalysis.com/p/long-live-the-short-king-why-4-hi) [ Long Live the Short King: Why 4-hi HBM Wins ](https://newsletter.semianalysis.com/p/long-live-the-short-king-why-4-hi) Myron Xie, Bryan Shan, and 3 others · Sep 13 [ Read full story ](https://newsletter.semianalysis.com/p/long-live-the-short-king-why-4-hi) Moreover, on B300 and week-0 stack, moving the Engram table back to HBM did not improve results and stay within run-to-run variance. This is a result of the work optimizing DRAM offload, such as async, and overlap. SSD Offloading On B200, we created a unoptimized vLLM fork and stored the Engram tables in memory-mapped files on local SSDs. File backing lets the OS reclaim table pages when other applications need RAM. Pages already cached in memory can be served without reading the SSD again. Furthermore, note that we were unable to turn on GDS The unoptimized SSD implementation changes how rows reach the GPU. It copies row IDs to the CPU, deduplicates them, gathers the requested rows into pinned buffers, copies those rows back to the GPU and dequantizes them. This work runs between segments of the GPU execution graph. Native UVA performs row selection and dequantization directly on the GPU, avoiding the CPU round trip. A warm filesystem cache removes physical SSD reads, but leaves the coordination, row gathering and transfers. This is why a file already cached in RAM can still perform worse than a pinned DRAM table. The comparison measures the whole serving path; it does not separate the time spent on each of these operations. B200 DRAM dominates both measured SSD serving curves in total tokens per dollar and P90 interactivity. Near 125 tokens/s/user, DRAM delivers 121 million total tokens per dollar versus 52 million for SSD. For production serving, SSD offloading is likely not worth the tradeoff. On the B200 configurations we measured, SSD offloading loses on both measures: every observed SSD point has a DRAM alternative that delivers higher P90 interactivity and more total tokens per dollar. The only points where Cheaper storage does not automatically produce a cheaper inference service. Moving Engram to SSD leaves the same four expensive GPUs and the rest of the server in place. Reclaiming RAM only creates an economic benefit if it enables a cheaper server configuration or additional useful capacity. The current unoptimized path provides neither benefit, and the filesystem cache still consumes RAM when table pages are resident. Next we will look at the mechanisms and specific implementation of ngrams from DeepSeekv4.1 Flash, LongCat, Qwen3.8 Flash Next.
Gears Of War: E-Day Defies Unreal Engine 5 Expectations, As 4K@60FPS Needs An RTX 5070 Ti Instead Of Flagship GPU
Gears of War: E-Day PC system requirements reveal tame Ultra specs for 4K 60 FPS with upscaling on Unreal Engine 5
AMD Plans 10% Price Hike Across GPUs, Chipsets, and Possibly CPUs
AMD is preparing for a series of price hikes in response to rising costs of acquiring silicon from its supply chain. According to ChannelGate, AMD is working on a 10% price increase across its products due to the upstream supply chain squeeze. Since most of AMD's silicon is manufactured by...
Updated PS6 Handheld Specs Leaked By KeplerL2: 24GB RAM, RDNA 5 GPU, PSSR 3.0
But pricing remains unknown.
Leaker Claims AMD And NVIDIA Delay Next-Gen GPUs To 2028, Except One RDNA 5-Based AT2 Chip Launching In 2027
AMD and NVIDIA may skip 2027 for next-gen GPUs, with only AMD's mid-range RDNA 5 AT2 launching next year alongside continued high prices.
Which Nvidia GPU Is Equivalent To The Meta Quest 3?
The Meta Quest 3 utilizes the Snapdragon XR2 Gen 2, which can rival an Nvidia GTX 960, but a one-to-one comparison doesn't paint the whole picture.
No One Can Believe The Current Price Of An Nvidia RTX 5090
The top-end GPU has hit astronomical prices, even in large retail stores
Which Nvidia GPU Is Equivalent To The Xbox Series S?
The 1650 SUPER from Nvidia's now discontinued GTX graphics card line comes closest to the GPU inside the Xbox Series S in terms of direct comparison.
NVIDIA Unveils RTX PRO 5500 "Blackwell" Workstation GPU with 84 GB GDDR7 Memory
NVIDIA today quietly added another card to its pro-viz product lineup with the RTX PRO 5500 "Blackwell" workstation GPU edition. Featuring 84 GB of GDDR7 memory, it is equipped with a GB202 die that includes 21,760 CUDA cores. This is the same CUDA core count as the consumer GeForce RTX 5090 SKU,...
DLSS Multi Frame Generation Has Now Been Modded Onto RTX 40, RTX 30 & Even RTX 20 GPUs
Modders unlock NVIDIA's DLSS Multi Frame Generation on RTX 20, 30, and 40 Series, expanding frame interpolation capabilities beyond the RTX 50 Series
Modders Unlock DLSS Multi Frame Generation For RTX 30-Series "Ampere" GPUs
After porting DLSS 5 to every recent NVIDIA generation, the modding community now seems to be turning its focus to Frame Generation. A new mod called DLSSG SM86 lets GeForce RTX 30-series owners run DLSS Multi Frame Generation in 2X and 4X modes, a feature NVIDIA officially reserves for RTX...
Apple announces A20 Pro chip with 2nm design and major performance gains
Apple today unveiled the A20 Pro, its first chip built on 2nm technology at its “Surprise and Shine” event, bringing major CPU, GPU, and AI upgrades to the iPhone 18 Pro lineup.
Mod Revives Multi-GPU Gaming With Dedicated DLSS 5 GPU for Impressive FPS Gain
If you were around during what many consider the golden age of PC gaming, back when SLI was still possible, and, going back even further, when gamers put dedicated PhysX cards into their PCs, the latest trick shown off by modder, Marcelo Guibout, on YouTube. Guibot showed off a gaming setup using...
DLSS 5 Mod Shifts Load to Second GPU, Doubling Frame-Rates and Latency
Plus, a second mod allows for sub-native rendering to boost performance.
User Demonstrates Offloading DLSS 5 Neural Rendering To The Second RTX 5060 Ti Using ReShade Add-On, Restoring Original Performance
DLSS 5 Neural Rendering can now run on a second GPU with MGPU Bridge, boosting performance in games like The Blood of Dawnwalker and Cyberpunk 2077.
User reports melted GeForce RTX 5090 power connector after testing DLSS 5 in NBA 2K27
Enabling DLSS 5 in NBA 2K27 pushes GPU power usage to its limit, and with that we've got our first post-DLSS 5 case of a 16-pin cable melting.
How to install Nvidia DLSS 5 on almost any RTX GPU in minutes
Your full guide to Nvidia's latest gaming upgrade.
Sony is preparing limited edition PlayStation 5 consoles in the style of GTA 6.
The gaming community is buzzing with discussions about possible limited edition versions of the PlayStation 5 and PlayStation 5 Pro consoles, timed to coincide with...
DLSS 5 was launched in Medal of Honor European Assault using the PCSX2 emulator on an RTX 4070 Super.
On Reddit, enthusiasts demonstrated the DLSS 5 upscaling technology in the classic shooter Medal of Honor: European Assault,...
Video card RTX 4090 showed a higher frame rate with DLSS 5 than the RTX 5080
A detailed comparative benchmarking of the NVIDIA GeForce RTX 5080 graphics accelerators with 16 GB of memory and GeForce has appeared online. RTX 4090 With...
HP’s latest OmniBooks get RTX Spark, 128 GB LPDDR5X RAM and 3K OLED displays
HP has unveiled the OmniBook Ultra 16 and OmniBook X 14, two RTX Spark-powered laptops that are among the first to offer the lower-tier 18-core N1X configuration with a 5,120-core Blackwell RTX GPU. Both come with up to 128 GB of LPDDR5X-9400 memory and 3K 120 Hz OLED displays.
DLSS 5 Will Be Coming to Older Graphics Cards After All
On Nvidia's forums, an Nvidia employee claimed that the company will be expanding DLSS 5 support to RTX 40 series graphics cards at some point, but we're probably going to be waiting a while for it.
ASUS Shows RTX Spark Mini PCs and Laptop Designs at IFA 2026
At IFA 2026 in Berlin, Germany, ASUS introduced mini PCs and laptops powered by NVIDIA's RTX Spark SoC. Starting with the GR1X mini, ASUS showcased the power of the RTX Spark N1X in grey or black aluminium enclosures. These are available with up to a 6,144-core GeForce RTX "Blackwell" GPU,...
AMD Threadripper Halo Station packs 96-core CPU and Instinct MI350P GPUs, up to 576GB GPU memory and 2TB system memory
AMD has shown its Threadripper Halo Station with a 96-core CPU, two Instinct MI350P GPUs and 2TB of system memory. The design is planned to support four GPUs with 576GB HBM3E.
NVIDIA Confirms DLSS 5 Support Extends To RTX 40 GPUs, But First Focusing On Model Optimization on RTX 50 Series
NVIDIA DLSS 5 is going to be released on older RTX 40 GPUs and will be available once the model is fully tuned on RTX 50 series.
Nvidia will officially bring DLSS 5 to older GPUs — but won’t give gamers full control
The cat is out of the bag.
NVIDIA RTX Spark PCs Arrive Next Month In Two “N1X” SoC Configs, One With a 20 Core CPU, 6144 Core GPU & Up To 128 GB Memory, Other With 18 Core CPU, 5120 Core GPU & Up To 32 GB Memory
NVIDIA's RTX Spark PCs are officially launching in October, powered by the N1X SoC, to deliver game-changing performance & efficiency.
Nvidia wants to turn your house of gaming PCs into an AI supercomputer
Nvidia's free RTX PAIR app spreads local AI workloads across multiple GPUs on your network, dramatically cutting down on processing times.
Leaker claims iPhone 18 Pro chip will have 7 GPU cores and faster RAM
A Chinese leaker with some track record has posted images of claimed circuit diagrams of the upcoming iPhone 18 Pro...
iPhone 18 Pro Chip Leak Indicates Major GPU and Memory Bandwidth Improvements
An alleged solder joint diagram for Apple's A20 Pro chip suggests the processor will deliver a number of significant upgrades, including a seven-core GPU, a wider memory interface, and a larger L2 cache for its efficiency cores. The claims come from a Weibo user who says the layout was reconstructed by mapping the chip's solder ball positions and overlaying them with die level images to sketch out where each block of the processor sits.
Apple’s A20 Pro Front Side Die Shot Shows New Record For GPU Core Count To Deliver Better Graphics Performance, New Packaging Allows For Increased Memory Bus Width
A leaked Apple A20 Pro die shot reveals a record GPU core count for upgraded graphics, while new WMCM packaging unlocks a wider 96-bit memory bus width
Early DLSS 5 Testing Suggests the RTX 5090's Single Power Connector Might Be the Bottleneck
Community testing of leaked DLSS 5 builds suggests the tech may already be hitting a wall on NVIDIA's own flagship GPU. Tom's Hardware ran the neural rendering model through Cyberpunk 2077, Hogwarts Legacy, and Control on an RTX 5090 Founders Edition, as well as an MSI RTX 5090 Lightning Z....
GPU prices are now 28% above MSRP on average, RTX 5090 explodes to 136% above MSRP
GPU prices have been tracked across 10 countries, revealing a troubling snapshot of the market that gamers have endured and will continue to endure.
DLSS 5 Officially Goes Live September 3 With NBA 2K27, Exclusive to RTX 50-series "Blackwell" GPUs
NVIDIA has confirmed that DLSS 5 will officially launch on September 3 alongside NBA 2K27, and it will remain exclusive to GeForce RTX 50-series GPUs. When asked directly at a Gamescom technical briefing whether older cards would ever receive support, NVIDIA gave a flat answer: "It's RTX...
NVIDIA DLSS 5 Performance Unveiled: 5x Faster Than Original Demo, 50% Perf Penalty on Average, But RTX 50 GPUs Offer “Hundreds of FPS” With MFG Enabled & More Optimizations On the Way
NVIDIA is sharing the first performance figures of its RTX 50 GPUs running DLSS 5, offering next-gen visuals at super-fast frame rates.
Nvidia’s controversial DLSS 5 arrives September 3rd and requires serious GPU horsepower
It seems to look better than the terrible initial examples.
Leaked DLSS 5 is Being Tested on RTX 20-series "Turing" GPUs, Now Runs on Emulators and Old DirectX Titles
The leaked DLSS 5 DLL keeps finding new homes, and support has now spread to RTX 20-series "Turing" GPUs. That architecture is three whole generations behind the NVIDIA-recommended "Blackwell," and modders are showing no signs of stopping either. Developer ShortFuse has released new DLSS-NR...
Xbox CEO Reacts To The Irony Of Rising Console Prices Being An Issue That Microsoft Itself Contributed To
Xbox CEO Asha Sharma believes it is "wonderful" that GPUs have use beyond gaming.
The Blood of Dawnwalker: Benchmarks of Graphics Cards and Processors
We completed large-scale benchmarks The Blood of Dawnwalker + GPU from RTX 3060 to RTX 5090 and RX 6750 XT – RX 9070 XT in conjunction with CPU AMD and Intel....
Public Confidence in One’s Own Healthcare Provider Remains High | The Annenberg Public Policy Center of the University of Pennsylvania
Americans’ confidence in their healthcare providers remains high while confidence in U.S. healthcare institutions has declined since the start of President Trump's second term, an APPC survey conducted from April 2021-August 2026 finds.
NVIDIA Shuts Down RTX 5000 Power Limit Control OC in Latest Driver Update
We recently reported on the open-source overclocking tool, mVolt+, unlocking power limit control on NVIDIA GeForce RTX 5000-series GPUs, allowing certain GPUs, like the RTX 5090, to draw up to 125 W of additional power and enabling significantly higher core clock speeds as a result. With the...
Pixel 11's Tensor G6 and Pixel 10's Tensor G5 still fall behind in performance per watt, testing shows
Geekerwan's first efficiency test of the Tensor G6 and G5 shows that both trail the Snapdragon 8 Elite and Apple's A-series chips in CPU and GPU performance per watt, while the Kirin 9030 Pro matches the Tensor chips despite being manufactured on a much older process node.
To play Fable at 4K resolution and 60 FPS, you will need a graphics card of the RTX 5080 level.
Playground Games has officially announced the system requirements for the role-playing game Fable for PC. The developers have prepared...
DLSS 5 Patched to Work on RTX 4000 "Ada Lovelace" GPUs Despite No Official Support
The modding community has been hard at work ever since the untimely release of DLSS-NR, a precursor to DLSS 5. Now, modders have gotten NVIDIA's leaked DLSS 5 tool working on RTX 4000 series cards, even though the tool was only built for RTX 5000 Blackwell hardware. The DLL leaked earlier this...
Nvidia earnings could reveal its fastest-growing business isn't chips: Chart of the Day
Wall Street is focused on Nvidia's GPUs, but the faster-growing story is sitting beside them.
Intel Crescent Island GPUs Pack Up To 32 Xe3P Cores, Optimized For Agentic AI With Low-Cost LPDDR5X That Reaches Up To 480 GB Capacity
Intel has given the rundown on its next-gen AI inference accelerator, Crescent Island, which packs 32 Xe3 cores and 480 GB of LPDDR5X memory.
AMD RDNA 4m Firmware Published For Linux Ahead Of Launch
In addition to AMD's Friday afternoon big code drop of all the UALink enablement patches for the Linux kernel and the AMDGPU driver, there was another big drop from AMD to end out the week..
Dev TFT: SPACE GODS Reviewed
Reflecting on SPACE GODS (and more) with the team behind and beyond the set!
Will the AI bubble break American politics?
Is the massive investment in artificial intelligence creating a catastrophic economic bubble? In this episode of "Make It Make Sense," Damir Marusic sits down with Impact Theory co-founder Tom Bilyeu to break down the risks of AI-driven debt, shadow banking, and fast-depreciating GPU hardware. They examine how an AI market correction could ignite political extremism, critique US fiscal policy under Donald Trump, and discuss what it will take to rebuild the American middle class.
GPU market evolves backwards as modded RTX 3080 20GB cards return as cheaper gaming alternatives
The RTX 3080 is nearly six years old, but 20GB modifications continue to appear in China. NVIDIA also continues supporting Ampere with newer DLSS models.