Investment bank Morgan Stanley has estimated that an astonishing 77% of all globally produced silicon wafers dedicated to AI accelerators will be consumed by none other than NVIDIA. Often, investment research by large investment banks like Morgan Stanley includes information from the semiconductor supply chain, which is constantly expanding to meet NVIDIA's demands. When looking at wafer volume for AI accelerators, it is estimated that in 2024, NVIDIA captured nearly 51% of wafer consumption for its chips, more than half of all demand. With NVIDIA's volume projected to grow to 77%, this represents more than a 50% year-over-year increase, which is incredible for a company of NVIDIA's size. Right now, NVIDIA is phasing out its H100 accelerators in favor of Blackwell 100/200 and the upcoming 300 series of GPUs paired with Grace CPUs.

NVIDIA is accelerating its product deployment timeline and investing a lot in its internal research and development. Morgan Stanley also projects that NVIDIA will invest almost $16 billion in its R&D budget, enough to endure four to five years of development cycles running three design teams sequentially and still delivering new products on an 18-24 month cadence. The scale of this efficiency and development rivals everyone else in the industry. With all this praise, NVIDIA's Q4 revenue report is coming in exactly a week on February 26, so we have to see what its CEO, Jensen Huang, will deliver and show some estimates for the coming months.

The days of 32-bit software support in NVIDIA's drivers are coming to an end, and with that, so does the support for the once iconic PhysX real-time physics engine. According to NVIDIA's engineers on GeForce forums, the lack of PhysX support has been quietly acknowledged, as NVIDIA's latest GeForce RTX 50 series of GPUs are phasing out support for 32-bit CUDA software, slowly transitioning the gaming world to the 64-bit software entirely. While older NVIDIA GPUs from the Maxwell through Ada generations will maintain 32-bit CUDA support, this update breaks backward compatibility for physics acceleration in legacy PC games on new GPUs. Users running these titles on RTX 50 series cards may need to rely on CPU-based PhysX processing, which could result in suboptimal performance compared to previous GPU generations.

A Reddit user reported frame rates dropping below 60 FPS in Borderlands 2 while using basic game mechanics with a 9800X3D CPU and RTX 5090 GPU, all because 32-bit CUDA application support on Blackwell architecture is depreciated. When another user booted up a 64-bit PhysX application, Batman Arkham Knight, PhysX worked perfectly, as expected. It is just that a massive list of older games, which gamers would sometimes prefer to play, is now running a lot slower on the most powerful consumer GPU due to the phase-out of 32-bit CUDA app support.

As reported almost a year ago, EMTEK graphics cards do not make regular appearances via Western reportage. The South Korean manufacturer has prepared its debut NVIDIA "Blackwell" GeForce RTX 50-series desktop product; as spotted by VideoCardz. The brand new EMTEK GeForce RTX 5080 MIRACLE WHITE D7 16 GB model has appeared online via Danawa's online price comparison engine. The lowest price comes in at 2,656,000 KRW (~$1841 USD), so there is clear evidence of major price hiking activities going on in South Korea. One store is offering EMTEK's cutting-edge white custom model for a cool 4,000,000 KRW (~$2773 USD).

EMTEK's GeForce RTX 5080 MIRACLE WHITE D7 16 GB graphics card conforms to Team Green's reference specifications; making the higher than expected entry fee even more painful to bear. We hope that prices level-off to reasonable lines in the near future. This snow-white custom design seemingly carries a couple of premium features; namely translucent ARGB-lit fans in a triple-formation. This integrated "dynamic lighting" system can be controlled through Windows 11 (via a USB-C connection), rather than proprietary software. VideoCardz reports that the South Korean graphics producer is not known to release its own RGB control suite. A cutaway render shows a very robust cooling solution that houses eight 6 mm heat pipes. Unfortunately, the PCB's shape is not as radical as the one found in an Ada Lovelace-based sibling. Judging from the internal promo shots, EMTEK has missed another opportunity to bung in a white board design, and attach similarly pale parts. Many NVIDIA AIBs should spend more time copying GALAX, Yeston, and PowerColor's (AMD-powered) homework—the latter's Spectral White Editions usually nail the almost all-white aesthetic.

Last Wednesday (February 12), an intriguing filing appeared online—the Eurasian Economic Commission (EEC) has received multiple registrations of NVIDIA GPUs; courtesy of OEM Global Technologies Ltd. The parent company of ZOTAC Technology Limited has registered a wide span of Team Green SKUs, going back many generations to 2014's GeForce GT 710 "Kepler 2.0" design. When ignoring everything south of the latest "Blackwell" GPU lineup, harukaze5719 noticed three lower-end additions. The rumor mill has NVIDIA launching its more budget-friendly GeForce RTX 5060 Ti and RTX 5060 (non-Ti) models at some point next month.

Board partner supply chain insiders believe that Team Green is clearing the current-gen flock; thus making way for lower-end "Blackwell" GPUs, but the newly-discovered registration of a GeForce RTX 5050 SKU has raised watchdog eyebrows. A rumored entry-level RTX 5050 Mobile variant turned up last December, but we have not heard much about a discrete desktop equivalent in 2025. NVIDIA had plans for a GeForce RTX 4050 model, but desktop "Ada Lovelace" AD107 GPU-based products did not emerge into existence. The latest ECC filing suggests that a next-gen successor could be on the way. Industry soothsayers predict a sub-$300 price point for alleged GeForce RTX 5050 8 GB graphics cards—potentially lined up for battle with Intel's current-gen B580 and B570 offerings.

NVIDIA and its memory partners, SK Hynix, Samsung, and Micron, are preparing a new memory form factor called System on Chip Advanced Memory Module—SOCAMM shortly. This technology, adapting to the now well-known CAMM memory module standard, aims to bring additional memory density to NVIDIA systems. Taking inspiration from NVIDIA's Project DIGITS, it has now been developed independently by NVIDIA outside of any official memory consortium like JEDEC. Utilizing a detachable module design, SOCAMM delivers superior specifications compared to existing solutions, featuring 694 I/O ports (versus LPCAMM's 644 and traditional DRAM's 260), direct LPDDR5X memory substrate integration, and a more compact form factor.

For reference, current-generation Project DIGITS is using NVIDIA GB10 Grace Blackwell Superchip capable of delivering one PetaFLOP of FP4 compute, paired with 128 GB of LPDDR5X memory. This configuration limits the size of AI models that run locally on the device, resulting in up to 200 billion parameter models running on a single Project DIGITS AI PC. Two stacked Project DIGITS PCs are needed for models like Llama 3.1 405B with 405 billion parameters. Most interestingly, memory capacity is the primary limiting factor; hence, NVIDIA devotes its time to developing a more memory-dense SOCAMM standard. Being a replacement compatible with LPDDR5, it can use the same controller silicon IP with only the SoC substrate being modified to fit the new memory. With NVIDIA rumored to enter the consumer PC market this year, we could also see an early implementation in consumer PC products, but the next-generation Project DIGITS is the primary target.

TrendForce's latest findings reveal that the NAND Flash market continues to be plagued by oversupply in the first quarter of 2025, leading to sustained price declines and financial strain for suppliers. However, TrendForce anticipates a significant improvement in the market's supply-demand balance in the second half of the year.

Key factors contributing to this shift include proactive production cuts by manufacturers, inventory reductions in the smartphone sector, and growing demand driven by AI and DeepSeek applications. These elements are expected to alleviate oversupply and support a price rebound for NAND Flash.

Yesterday, NVIDIA confirmed that board partner GeForce RTX 5070 Ti graphics cards (minus a Founders Edition) will be hitting global retail markets on February 20. Team Green has set a baseline MSRP of $749 (USD) for North America, and €884 (plus region variable VAT) for European territories. Recent accidental e-tail listings have presented daunting price points (in Austria and France), that sit far higher than NVIDIA's guideline figure. A fresh VideoCardz news report piles on extra pain; various readership tip-offs have indicated that next week's population of launch products will be composed of mostly very expensive offerings. CaseKing.de has listed a grand total of twelve ASUS, GIGABYTE and MSI GeForce RTX 5070 Ti cards—the cheapest non-overclocked option comes in at €1149 (including customary 19% sales tax), plus shipping costs starting at €4.99. GIGABYTE's premium-tier AORUS RTX 5070 Ti MASTER is priced at €1399 (incl. VAT). This German retail outlet is not known to implement generous discounts, especially early on in a product's lifespan. VideoCardz looked elsewhere for evidence of baseline MSRP conformant AIB cards. Press outlets could be analyzing placeholder numbers, so things could change closer to release day.

Proshop in Denmark and Finland seem to be the only e-tail outlets (in Europe/Scandinavia) that offer a model that conforms to official NVIDIA minimum pricing decree. The non-overclocked barebones Inno3D RTX 5070 Ti X3 card is priced at 6899 kr (incl. VAT) in Denmark, and €924 (incl. VAT) in Finland. An extra €75 premium grants access to an overclocked sibling: Inno3D RTX 5070 Ti X3 OC. Potential buyers—in North America—have one baseline MSRP-friendly option, courtesy of Micro Center. VideoCardz highlighted the PNY RTX 5070 Ti Overclocked Triple Fan model's price point of $750. At the time of writing, they believe that this is the only example of a factory-overclocked card sticking to Team Green guidelines. AMD's incoming Radeon RX 9070 GPU series is set to battle it out with NVIDIA's next wave of "Blackwell" GPUs. According to industry whispers, Team Red will be deeply involved in analysis of GeForce RTX 5070 Ti market performance. Speculative first wave RDNA 4-based graphics card pricing has leaked online this week; will they commit to undercutting their main competition?

Retailers are set to open up pre-order floodgates for upcoming GeForce RTX 50-series laptops on February 25, as we learned earlier this week. According to a new DigiTimes report, the launch of mobile devices—sporting Team Green "Blackwell" GPUs—is expected to be "significantly delayed." A loose March launch window was teased during Jensen Huang's keynote presentation at CES 2025, but supply chain insiders have claimed that high-end RTX 50 laptops were "originally planned to be launched in January 2025." Additionally, they surmise that mid-range and low-end offerings are postponed to April. DigiTimes believes that the rumored postponements have surprised supply chain moles; Team Green is not known to delay product launches. Extenuating circumstances are cited as the reason behind alleged deferred release windows, but insiders have not yet determined the extent of lengthened launch parameters.

An anonymous source stated: "NVIDIA, which has never been late in the past, also encountered this situation. It is probably related to NVIDIA's full sprint to AI servers. Even though there are differences in server and PC chip design and manufacturing processes, the company's resource allocation may still affect the debugging efficiency of new products." Other insiders have murmured about GeForce RTX 50-series mobile GPU performance not meeting expectations. Rumors have swirled about problems with early sample units; most notably the encountering of major screen issues when the "hardware is turned on." Laptop/notebook supply chain insiders reckon that manufacturers have anticipated a healthy level of growth in 2025—thanks to the emergence of new NVIDIA graphics cards—but targets have been reduced, due to anticipated delays. Optimistic industry chatter predicts higher education students and esport enthusiasts driving unit sales upward, following a stagnant 2024 market.

Hewlett Packard Enterprise announced today that it has shipped its first NVIDIA Blackwell family-based solution, the NVIDIA GB200 NVL72. This rack-scale system by HPE is designed to help service providers and large enterprises quickly deploy very large, complex AI clusters with advanced, direct liquid cooling solutions to optimize efficiency and performance. "AI service providers and large enterprise model builders are under tremendous pressure to offer scalability, extreme performance, and fast time-to-deployment," said Trish Damkroger, senior vice president and general manager of HPC & AI Infrastructure Solutions, HPE. "As builders of the world's top three fastest systems with direct liquid cooling, HPE offers customers lower cost per token training and best-in-class performance with industry-leading services expertise."

The NVIDIA GB200 NVL72 features shared-memory, low-latency architecture with the latest GPU technology designed for extremely large AI models of over a trillion parameters, in one memory space. GB200 NVL72 offers seamless integration of NVIDIA CPUs, GPUs, compute and switch trays, networking, and software, bringing together extreme performance to address heavily parallelizable workloads, like generative AI (GenAI) model training and inferencing, along with NVIDIA software applications. "Engineers, scientists and researchers need cutting-edge liquid cooling technology to keep up with increasing power and compute requirements," said Bob Pette, vice president of enterprise platforms at NVIDIA. "Building on continued collaboration between HPE and NVIDIA, HPE's first shipment of NVIDIA GB200 NVL72 will help service providers and large enterprises efficiently build, deploy and scale large AI clusters."

NVIDIA's official GeForce Twitter handle just confirmed that the GeForce RTX 5070 Ti will be available from February 20, 2025. This is the third RTX 50-series "Blackwell" graphics card launch. The RTX 5070 Ti is positioned in the enthusiast segment considering its starting price of $749, with custom-design OC cards expected to be priced well above $800. Although officially being recommended for maxed out gameplay at WQHD (2560 x 1440) or ultrawide 3440 x 1440 resolutions, we expect the RTX 5070 Ti to be capable of gameplay at 4K UHD considering its predecessors, the RTX 4070 Ti and RTX 3070 Ti, each were at launch.

The GeForce RTX 5070 Ti is carved out from the same 5 nm "GB203" silicon that powers the RTX 5080 launched late last month. It is endowed with 70 out of 84 SM present on the silicon, which works out to 8,960 CUDA cores, 280 Tensor cores, 70 RT cores, 280 TMUs, and possibly 96 ROPs. The card gets 16 GB of [at least] 28 Gbps GDDR7 memory, giving it 896 GB/s of memory bandwidth on tap, a massive increase from the 12 GB and 504 GB/s of its predecessor. Then there are the Blackwell-exclusive technologies such as DLSS 4 Multi Frame Generation, Neural Rendering, Mega Geometry for ray tracing, and an updated display engine with support for DisplayPort 2.1b UHBR20. The picture NVIDIA used in its tweet shows an abstract-looking card, which could be a sign that there won't be a Founders Edition card of the RTX 5070 Ti.

ASUS has released version one of a BIOS update tool for its ultra-premium air-cooled ROG Astral graphics cards, OC and standard flavors. Yesterday's update advertises an improved "Quiet Mode," that implements a "more silent fan curve." TechPowerUp reviewed the quad-fan configured ASUS ROG Astral GeForce RTX 5090 and RTX 5080 OC Edition cards late last month. Resident GPU evaluator, W1zzard, awarded the astronomically-priced flagship model with an "Editor's Choice" badge, but simultaneously pinned on a "But Expensive" honor. The lesser (GB203-based) Astral did not receive any accolades. Both models have courted criticism for louder than expected operation; W1zzard did not enjoy listening to the Astral GeForce RTX 5090 OC Edition's collective fan concerto. "Out of the box" settings were not great, and the quiet BIOS mode did not "help much" in reducing the highest-end Astral's cacophony.

TPU's GPU guru elaborated further: "the second BIOS runs a more relaxed fan curve, but it's not much quieter and achieves 36.5 dBA with 70°C. Sure, good temperature, but isn't the point of a 'quiet' BIOS that isn't quiet, even if temperatures are higher? MSI's Suprim lineup does much better noise-wise, with temperatures that aren't that much higher." The ASUS ROG Astral RTX 5080 OC Edition seems to be a less noisy prospect, given that its cooling solution is not dealing with the largest "Blackwell" GPU die. Its sound signature was less offensive, but W1zzard reckoned that there was room for improvement. On this subject he stated: "with the default (performance) BIOS, temperatures are extremely low, but noise levels are a little bit on the high side with 36 dBA. I would have preferred a more balanced setting. Good thing that ASUS includes a secondary quiet BIOS with their card. Now the card runs whisper quiet, emitting only 26 dBA, which is highly impressive for a card in this performance segment. It is not the quietest card though, the MSI Suprim SOC is a tiny bit quieter, and it is so out of the box, without requiring a manual BIOS switch change." ASUS has seemingly absorbed initial feedback from review outlets (plus early adopters)—yesterday's update arrived just under two weeks from launch time. Watch out for possible upcoming reassessments.

European PC hardware retailers are readying themselves with NVIDIA GeForce RTX 5070 Ti partner models, ahead of the GPU's freshly scheduled launch on February 20. Tech news sites and industry watchdogs have already sniffed out preliminary pricing; courtesy of accidental listings of MSI custom card designs in France and Austria. We are more than a week away from GeForce RTX 5070 Ti release day, so speculative price points are subject to change. The Austrian arm of MediaMarkt was quick to correct matters after an accidental leak, and proceeded to remove an MSI GeForce RTX 5070 Ti GAMING TRIO OC PLUS product page from its webstore. VideoCardz managed to preserve fundamental details, prior to MediaMarkt AT's de-listing of the offending item. The card's alleged €1169 asking price has raised Central European eyebrows—factoring in the region's customary VAT, we are looking at an approximate 32% increase over NVIDIA's baseline MSRP (€884). It should be noted that a Founders Edition does not exist in Team Green's upper-mid "Blackwell" GPU segment.

MSI's price point for its GeForce RTX 5070 Ti VENTUS 3X OC PLUS card is expected to reside closer to Team Green's recommendation, but this barebones offering sits within the company's "entry-level" product tier. The manufacturer's GAMING TRIO OC PLUS range tends to slot right into middle of product hierarchies—provisioned RGB lighting, larger heatsinks and miscellaneous fancy features often demand an upcharge at retail. Tomasz Gawroński (aka GawroskiT) happened upon additional MSI listings in France; an unnamed e-tailer has priced the GeForce RTX 5070 Ti GAMING TRIO OC PLUS at €1149 (€20 cheaper than MediaMarkt AT). The upper crust MSI GeForce RTX 5070 Ti VANGUARD SOC LIMITED EDITION model demands that French customers part way with €1199, for the pleasure of ownership.

The new ASUS ROG Astral GeForce RTX 5090 and RTX 5080 designs have attracted a lot of attention, due to an unusual cooling configuration that includes a backplate-mounted fourth fan. At CES 2025, MSI teased onlookers with a placard adorned with a GeForce RTX 32G "Lightning" Special Edition model—featuring a "FiveFrozr" cooling solution. A traditional triple-fan setup is placed in the expected shroud location, but two additional units are integrated into the card's backplate. According to recent reports and early reviews, GIGABYTE has deployed a somewhat related system, albeit entirely optional (depending on user discretion). The Taiwanese manufacturer sent its AORUS RTX 5090 and 5080 MASTER models to market last week. These premium card designs feature the company's new "Screen Cooling Plus" system. CES press material claims that the "extra air-boosting fan" grants more airflow.

GIGABYTE's fourth fan has flown under the radar, but major hardware news outlets have just picked up on initial impressions. Singapore's HardwareZone appreciated the inclusion of an optional extra—with their AORUS RTX 5080 MASTER sample—but criticized GIGABYTE's slightly undercooked implementation. Their reviewer did not evaluate whether the modular part made any difference in terms of reducing temperatures—instead, they opined: "to further improve cooling, the card also comes bundled with a separate 120 mm RGB fan that you can place on the back of the card to pull air out—a design reminiscent of the ROG Astral RTX 5080's built-in cooling solution. It's a practical touch but not an elegant one, as it means having to deal with additional cables to tidy up since—oddly enough—the card itself does not come with a power connector for the extra fan." GLITCHED.online, a South African tech site, took GIGABYTE's AORUS RTX 5080 MASTER ICE card for a test drive—they found that the extra bit of cooling potential made a difference, but it was "almost unnoticeable." We hope that GIGABYTE will send review samples to TPU's W1zzard in the near future. Will the fourth fan make any difference on the AORUS RTX 5090 MASTER model?

Earlier today, NVIDIA happily declared that: "GeForce RTX 50-series laptop pre-orders start February 25 from OEMs." A March launch window was mentioned during the company's official unveiling of its "Blackwell" GPU architecture at CES 2025. Mobile variants of the GeForce RTX 5090, RTX 5080, and RTX 5070 Ti are expected to ship with laptops next month. GeForce RTX 5070 Mobile-equipped devices are due in April. As reported by VideoCardz, a Finnish retailer's webstore has confirmed the February 25 pre-order start date. The Gigantti.fi site lists several new ASUS ROG Zephyrus and Strix laptops; configured with the latest Intel Core Ultra 200HX or AMD "Strix Point/Fire Range" mobile APUs, and NVIDIA GeForce RTX 50-series discrete graphics solutions. The most expensive option is listed with a €5499 price point—interested buyers are looking at a spec that includes a 16-inch OLED screen, Core Ultra 9-285H processor, 32 GB LPDDR5X RAM, 1 TB M.2 PCIe SSD, and a GeForce RTX 5090 32 GB mobile GPU. At the time of writing, Gigantti's "cheapest" option is a €2999 ROG Strix 16 model; sporting a Ryzen 9 9955HX APU and GeForce RTX 5070 Ti 12 GB Mobile GPU.

NVIDIA disclosed official MSRPs during last month's CES presentation: $2199 for RTX 5090, $2199 for RTX 5080, $1599 for RTX 5070 Ti, and $1299 for the RTX 5070. Western press outlets have scouted North American online retail outlets. Tom's Hardware observed (ASUS and HP) price ranges starting at $1800. The most expensive offering came in at $4200; an ultra high-end ROG Strix Scar 18 laptop featuring a GeForce RTX 5090 Mobile GPU and an Intel Core Ultra 9 275HX CPU. Team Green's upcoming launch partners include Acer, ASUS, Dell, Gigabyte, HP, Lenovo, Mechrevo, MSI and Razer. NVIDIA seems to have extra GeForce RTX 50-series announcements in the pipeline—they recommend that potential buyers: "stay tuned for more details!"

According to early February reportage, Team Green's GeForce RTX 5060 and RTX 5060 Ti graphics cards are expected to launch onto market next month. Very basic technical information has leaked online; insiders reckon that both product tiers will be utilizing the NVIDIA "Blackwell" GB206 GPU. Rumors have swirled regarding intended VRAM configurations—loose online declarations point to variants being prepared with 8 or 16 GB of GDDR7 VRAM, on a 128-bit bus. Regulatory filings indicate two different configs with the eventual arrival of GeForce RTX 5060 Ti models, but certain industry watchdogs insist that the GeForce RTX 5060 SKU will be an 8 GB-only product.

A curious-looking ZOTAC Trinity OC White Edition GeForce RTX 5060 16 GB variant surfaced via a TikTok video—post-analysis, expert eyes declared that the upload contained doctored material. A BenchLife.info report pointed to a notable inconsistency on the offending item's retail packaging: "DLSS 3 should not appear on the GeForce RTX 50 series box, because the Blackwell GPU architecture focuses on DLSS 4." The publication presented evidence of ZOTAC RTX 4070 Ti SUPER Trinity OC White Edition box art being repurposed in the TikToker's video. Hardware soothsayer MEGAsizeGPU added their two cents: "this is fake. There is no plan for a GeForce RTX 5060 16 GB, and the box is photoshopped from the last-gen ZOTAC box." At the end of their report, BenchLife reckons that NVIDIA has not sent a "GeForce RTX 5060 color box template" to its board partners.

Unlucky owners of problematic GeForce RTX 5090 and RTX 5080 graphics cards have posted feedback across various online community sites. Press outlets started to take notice of these documented issues soon after the launch of NVIDIA's debut wave of "Blackwell" GPUs. PC Gamer has "kept track" of reports relating to black screens and miscellaneous failures—the site published an investigative article late last week, following user feedback "hitting critical mass" across Reddit and Team Green's own forum. A request for comment was sent over to NVIDIA HQ; PC Gamer received a brief response. A company spokesperson confirmed that their team is: "investigating the reported issues with the RTX 50-series."

Several PC hardware community members have documented their troubleshooting experiences—the most common suggestion involves downshifting from a PCIe 5.0 connection to 4.0, on the PEG-16 graphics port. Unfortunately, this step did not resolve black screen issues for certain owners—a member of the buildapc subreddit explored a wide array of troubleshooting channels. They re-installed Windows 11 (23H2), adjusted BIOS settings, experimented with monitor connections, and played around with drivers. Best results were produced by connecting a single monitor to their MSI GeForce RTX 5090 GAMING TRIO OC's DisplayPort, with nothing else hooked up to the other ports (DP and HDMI). They suspect that Team Green's GPU drivers could be the source of frustrations; corroborated by a recent VideoCardz news piece. In addition, the 572.16 driver is reportedly affecting "certain GeForce RTX 40-series." A smaller number of owners have discussed a "bricking" of cards; VideoCardz believes that China-exclusive GeForce RTX 5090D models are suffering the most. Manli will be analyzing a "bricked" unit at its service center, in the near future. Colorful did not reply with a comment on the situation.

According to ITHome, AMD is preparing to disrupt its competition with aggressive pricing for its upcoming RX 9000 series. The RX 9070 XT, built on the RDNA 4 architecture, is expected to launch at $599, positioning it directly against NVIDIA's RTX 5070 Ti, which carries a $749 price tag. With this competitive pricing, AMD aims to revitalize its market position following lower-than-expected sales of the RX 7000 series, causing it to lose some market share. The upcoming RX 9070 XT features the Navi 48 core running at 2.97 GHz, complemented by 16 GB of GDDR6 memory across a 256-bit bus. Architecture's enhanced AI upscaling capabilities, already demonstrated in the PlayStation 5 Pro, could offer compelling performance advantages over current-generation cards. The base RX 9070 model is anticipated to debut at $499, creating a focused attack on multiple market segments, including NVIDIA's RTX 5070, priced at $549.

AMD reportedly plans to accelerate the end-of-life timeline for its RX 7800 XT, currently priced at $479. Sources from IT Home suggest production ceased as early as January, months ahead of the planned initial third-quarter 2025 termination. This accelerated timeline suggests AMD's confidence in the RX 9000 series' ability to deliver superior price-to-performance metrics. The March 2025 launch window for the RX 9000 series arrives at a critical point in the GPU market, as NVIDIA rolls out its Blackwell-based RTX 50 series. AMD's aggressive pricing strategy and the architectural improvements in RDNA 4 positions the company to challenge NVIDIA's market dominance, at least in the $500-$600 price range. This competitive positioning could trigger NVIDIA price adjustments, potentially benefiting consumers who have faced consistently high GPU prices in recent years.

MSI France posted a countdown timer on its regional website for the GeForce RTX 5070 Ti, which winds down to February 20, 2025, 3 PM CET. This is our first confirmation that NVIDIA's third graphics card from the RTX 50-series "Blackwell" will be launching next week. MSI already has product pages of its custom design RTX 5070 Ti and RTX 5070 graphics cards up on its website. The RTX 5070 Ti lineup will be topped by the RTX 5070 Ti Vanguard series, followed by the RTX 5070 Ti Gaming Trio series. There's also the RTX 5070 Ti Inspire 3X series targeting creators. There are two "entry level" SKUs, the RTX 5070 Ti Ventus 3X, which the company will price close to the NVIDIA baseline, and the RTX 5070 Ti Shadow 3X, which meets NVIDIA SFF-Ready requirements.

MSI has, quite appropriately, introduced new SHADOW 3X graphics card models in a low-key manner. These stealth-black GeForce RTX 5080 and RTX 5070 Ti custom designs have just appeared online via official product pages. We expect to see press material published in the near future, perhaps closer to the launch of NVIDIA's next wave of Blackwell GPUs. MSI's newly discovered product pages place emphasis on the SHADOW 3X's essential features: "a performance-focused design that delivers the gaming experience players want, making it the ideal choice when upgrading or building a gaming rig." VideoCardz believes that the incoming SHADOW 3X cards will be the first to reach Western markets—previous-gen models have released as Chinese regional exclusives. The GeForce RTX 4070 Ti SUPER 16G SHADOW 3X overclocked (OC) and non-OC SKUs appear to re-use MSI's familiar VENTUS 3X design, just tinted many shades darker.

The new SHADOW 3X cards seem to utilize a fresh design, with no recycling from a similarly appointed sibling product family. MSI's latest barebones/essential tier brings a minimal set of perks: a sizable heatsink, a triple-fan cooling solution, a single 16-pin power connector and a straightforward shroud and backplate aesthetic. SHADOW 3X eschews anything RGB lighting-related. Potential buyers will not be dealing with fancy vapor chambers or liquid cooling systems. MSI will be offering overclocked editions, as well as standard models that conform to NVIDIA reference specifications. TechPowerUp's well-maintained GPU database will be updated with these new additions.

AI reasoning models and agents are set to transform industries, but delivering their full potential at scale requires massive compute and optimized software. The "reasoning" process involves multiple models, generating many additional tokens, and demands infrastructure with a combination of high-speed communication, memory and compute to ensure real-time, high-quality results. To meet this demand, CoreWeave has launched NVIDIA GB200 NVL72-based instances, becoming the first cloud service provider to make the NVIDIA Blackwell platform generally available. With rack-scale NVIDIA NVLink across 72 NVIDIA Blackwell GPUs and 36 NVIDIA Grace CPUs, scaling to up to 110,000 GPUs with NVIDIA Quantum-2 InfiniBand networking, these instances provide the scale and performance needed to build and deploy the next generation of AI reasoning models and agents.

NVIDIA GB200 NVL72 on CoreWeave
NVIDIA GB200 NVL72 is a liquid-cooled, rack-scale solution with a 72-GPU NVLink domain, which enables the six dozen GPUs to act as a single massive GPU. NVIDIA Blackwell features many technological breakthroughs that accelerate inference token generation, boosting performance while reducing service costs. For example, fifth-generation NVLink enables 130 TB/s of GPU bandwidth in one 72-GPU NVLink domain, and the second-generation Transformer Engine enables FP4 for faster AI performance while maintaining high accuracy. CoreWeave's portfolio of managed cloud services is purpose-built for Blackwell. CoreWeave Kubernetes Service optimizes workload orchestration by exposing NVLink domain IDs, ensuring efficient scheduling within the same rack. Slurm on Kubernetes (SUNK) supports the topology block plug-in, enabling intelligent workload distribution across GB200 NVL72 racks. In addition, CoreWeave's Observability Platform provides real-time insights into NVLink performance, GPU utilization and temperatures.

Supply chain and board partner (AIB) insiders have once again signalled a murky future for Team Green's "Ada Lovelace" generation of gaming graphics cards—Chinese industry soothsayers believe that supplies of GeForce RTX 4060 GPUs will change significantly throughout February. A somewhat similar disclosure popped up online half-way through January—when members of the Board Channels discussion board predicted an end of the month stock depletion of NVIDIA GeForce RTX 4070 GPUs. Several web entities believe that NVIDIA is paving the way for a next wave of GeForce RTX 50-series "Blackwell" graphics cards.

Gory details appeared on Board Channels—yesterday morning's China market-focused bulletin stated: "supply of the main models of the RTX 4060 series will be greatly reduced from the first quarter of 2025, that is, from February 2025 in the domestic market, NVIDIA's supply of RTX4060 series GPUs will be greatly reduced, which is at least 60% less than the estimated Q4 of 2024." Team Green and its partners are reportedly working on a March launch window for GeForce RTX 5060 (non-Ti) and RTX 5060 Ti graphics card models—neatly aligning with the RTX 4060-related timeframes proposed by industry insiders. The Board Channels article added further clarification: "the number of GPUs that each core AIC brand manufacturer can get from NVIDIA will be greatly reduced."

The NVIDIA RTX 50-series witnessed quite a contentious launch, to say the least. Hindered by abysmal availability, controversial generational improvement, and whacky marketing tactics by Team Green, it would be safe to say a lot of passionate gamers were left utterly disappointed. That said, while the desktop cards have been the talk of the town as of late, the RTX 50 Laptop counterparts are yet to make headlines. Occasional leaks do appear on the interwebs, the latest one of which seems to indicate the 3D Mark Time Spy performance for the RTX 5080 Laptop GPU. And the results are - well, debatable.

We do know that the RTX 5080 Laptop GPU will feature 7680 CUDA cores, a shockingly modest increase over its predecessor. Considering that we did not get a node shrink this time around, the architectural improvements appear to be rather minimal, going by the tests conducted so far. Of course, the biggest boost in performance will likely be afforded by GDDR7 memory, utilizing a 256-bit bus, compared its predecessor's GDDR6 memory on a 192-bit bus. In 3D Mark's Time Spy DX12 test, which is somewhat of an outdated benchmark, the RTX 5080 Laptop managed roughly around 21,900 points. The RTX 4080 Laptop, on an average, rakes in around 18,200 points, putting the RTX 5080 Laptop ahead by almost 19%. The RTX 4090 Laptop is also left behind, by around 5%.

Acer today announced the expansion of its Predator Helios Neo portfolio with the all-new Predator Helios Neo 16 AI and Predator Helios Neo 18 AI gaming laptops. These cutting-edge devices are powered by up to an Intel Core Ultra 9 275HX processor and up to an NVIDIA GeForce RTX 5070 Ti Laptop GPU, built to deliver game-changing AI capabilities for players entering the next era of gaming with these accessible, power-driven devices.

Powered by NVIDIA Blackwell, NVIDIA GeForce RTX 50 Series Laptop GPUs bring game-changing capabilities to gamers and creators. Equipped with a massive level of AI horsepower, the RTX 50 Series enables new experiences and next-level graphics fidelity. Multiply performance with NVIDIA DLSS 4, generate images at unprecedented speed, and unleash creativity with NVIDIA Studio. Plus, access NVIDIA NIM microservices - state-of-the-art AI models that let enthusiasts and developers build AI assistants, agents, and workflows with peak performance on NIM-ready systems.

As documented on the Linus Tech Tips (LTT) subreddit, a lucky customer presented their freshly-delivered Founders Edition (FE) GeForce RTX 5080 graphics card model. Newly registered member, EssDee3D, had much to celebrate—having acquired launch stock, presumably at a non-scalped price—but they were confused by their package's contents. A question was pushed out to the LTT community: "can someone explain what happened here? Direct from NVIDIA...I ordered an RTX 5080, and I got a graphics card with 5090 engraved on it. The outer box has the 5080 SKU on it. Wondering if anyone else has seen something like this before?" Unique circumstances—possibly caused by a mix-up during the manufacturing process—have produced an oddball hybrid. EssDee3D proceeded to add this curiosity to their PC build—following an absorption of (very mixed) feedback from other LTT members. Basic diagnostics—performed in a Windows OS environment—revealed that the card in question housed a bog-standard GeForce RTX 5080 GB203-based GPU.

NVIDIA's two Founders Edition shroud designs—for its GeForce RTX 5090 and 5080 SKUs—look nigh identical. TechPowerUp's resident graphics card reviewer extraordinaire—W1zzard—captured a handful of comparison shots for usage in his evaluation of NVIDIA's GeForce RTX 5080 Founders Edition model. The GeForce RTX 5090 FE's cooling solution—utilizing liquid metal, instead of thermal paste—is more robust when compared to the one present on its step-down sibling—rated for a TDP of 575 W (versus 360 W). Evidently, designed to temper any radiance emitted by Team Green's GB202 GPU. EssDee3D's "chimera" card was placed in the correct packaging; a 5080-labelled paper-fiber box—but the shroud's backside advertises itself as a GeForce RTX 5090 FE. PC hardware news outlets and Redditors are wondering whether additional examples—of jumbled up "Blackwell" GPU Founders Edition parts—will appear online over the next month or two, or three...or more.

The next generation of graphics performance has arrived. We've prepared an all-new series of cards: ROG Astral. Featuring a new, sophisticated design and an outstanding cooling solution, the ROG Astral GeForce RTX 5090 and ROG Astral GeForce RTX 5080 are your premium picks for supercharging the performance of your gaming PC. All this new hardware in the ROG Astral GeForce RTX 5090 requires no small amount of power so that it can stretch its legs and run. Your PSU should be capable of at least 1000 W to run this card—more on that later. The circuitry that delivers this power is just as important, and it's one reason why many enthusiasts prefer ROG graphics cards. We've equipped the ROG Astral GeForce RTX 5090 and 5080 for premium power delivery with 80-amp MOSFETs that can supply over 35% more headroom than standard designs. A massive 24-phase VRM array for the GPU and a seven-phase VRM for the GDDR7 memory chips distribute the work of supplying power, ensuring rock-solid stability and long-lasting performance. To give you peace of mind that your 16-pin PCIe power connector is seated properly, we provide monitoring through Power Detector+ in the GPU Tweak III app so that you can verify that the connector is fully seated. The app can even tell you exactly which pin is not seated properly, if that ever becomes a concern.

Ada, meet Blackwell
With the GeForce RTX 50 Series, NVIDIA debuts its latest Blackwell architecture. Armed with fifth-gen Tensor cores, new streaming multiprocessors optimized for neural shaders, and fourth-gen Ray Tracing cores built for Mega Geometry, the new graphics cards unlock access to the next generation of graphics technologies. For many gamers, the highlight of the new architecture is DLSS 4. DLSS is a revolutionary suite of neural rendering technologies that uses AI to boost FPS, reduce latency, and improve image quality. The latest breakthrough, DLSS 4, brings new Multi Frame Generation and enhanced Ray Reconstruction and Super Resolution. But there's more. NVIDIA Reflex 2 with Frame Warp provides game-winning responsiveness, and these cards are equipped to give you the best experience with ray-traced graphics yet.