Instructions per clock is a metric used to define and compare CPU architecture performance usually. However, enthusiast colleagues at ComputerBase had an idea to test the IPC improvement in GPUs, comparing it across current and past generations. NVIDIA's Blackwell-based GeForce RTX 50 series faces off against the Ada Lovelace-based RTX 40 generation, while AMD's RDNA 4-powered Radeon RX 9000 lineup challenges the RDNA 3-based RX 7000 series. For NVIDIA, the test used RTX 5070 Ti and 4070 Ti SUPER, aligning ALU counts and clock speeds and treating memory bandwidth differences as negligible. For AMD, the test matched the RX 9060 XT to the RX 7600 XT, both featuring identical ALUs and GDDR6 memory. By closely matching shader counts and normalizing for clock variations, ComputerBase isolates IPC improvements from other hardware enhancements. In rasterized rendering tests across 19 popular titles, NVIDIA's Blackwell architecture delivered an average IPC advantage of just 1% over the older Ada Lovelace.

This difference could easily be attributed to normal benchmark variance. Ray tracing and path tracing benchmarks showed no significant IPC uplift, leaving the latest generation essentially on par with its predecessor when normalized for clock and unit count. AMD's RDNA 4, by contrast, exhibited a substantial IPC leap. Rasterized performance improved by around 20% compared to RDNA 3, while ray-traced workloads enjoyed a roughly 31% gain. Path tracing results were even more extreme, with RDNA 4 delivering nearly twice the FPS, a 100% increase over its predecessor. These findings suggest that NVIDIA's performance improvements primarily stem from higher clock speeds, increased execution unit counts, and enhanced features. AMD's RDNA 4 represents a significant architectural advance, marking its most notable IPC gain since the original RDNA launch.

A research team from Coburg University of Applied Sciences and Arts in Germany, alongside AMD Germany, introduced a game-changing approach to procedural tree creation that runs entirely on the GPU, delivering both speed and flexibility, unlike anything we've seen before. Showcased at High-Performance Graphics 2025 in Copenhagen, the new pipeline utilizes DirectX 12 work graphs and mesh nodes to construct detailed tree models on the fly, without any CPU muscle. Artists and developers can tweak more than 150 parameters, everything from seasonal leaf color shifts and branch pruning styles to complex animations and automatic level-of-detail adjustments, all in real-time. When tested on an AMD Radeon RX 7900 XTX, the system generated and pushed unique tree geometries into the geometry buffer in just over three milliseconds. It then automatically tunes detail levels to maintain a target frame rate, effortlessly demonstrating stable 120 FPS under heavy workloads.

Wind effects and environmental interactions update seamlessly, and the CPU's only job is to fill a small set of constants (camera matrices, timestamps, and so on) before dispatching a single work graph. There's no need for continuous host-device chatter or asset streaming, which simplifies integration into existing engines. Perhaps the most eye-opening result is how little memory the transient data consumes. A traditional buffer-heavy approach might need tens of GB, but researcher's demo holds onto just 51 KB of persistent state per frame—a mind-boggling 99.9999% reduction compared to conventional methods. A scratch buffer of up to 1.5 GB is allocated for work-graph execution, though actual usage varies by GPU driver and can be released or reused afterward. Static assets, such as meshes and textures, remain unaffected, leaving future opportunities for neural compression or procedural texturing to further enhance memory savings.

Colorful Technology Company Limited, a leading brand in gaming PC components, gaming laptops, and HiFi audio products, proudly introduces the SMART 900 AI Mini PC - a powerful, compact system designed to accelerate AI-driven content creation. Engineered for students and professionals, the SMART 900 combines cutting-edge performance with a sleek industrial design, pushing the boundaries of what a mini-PC can deliver.

At its core is the new Ryzen AI Max+ 395 processor, built on the advanced Zen 5 architecture. With 16 cores and 32 threads, a turbo frequency of up to 5.1 GHz, and 80 MB of cache, it delivers desktop-class performance in a 4-liter form factor. The integrated Radeon 8060S graphics, based on the RDNA 3.5 architecture with 40 compute units, offer GPU performance rivaling discrete mobile GPUs. For AI tasks, a dedicated 50 TOPS NPU provides outstanding acceleration for machine learning and generative workloads. The SMART 900 is built for creators on the move. It features a rich set of connectivity options including 3x USB 3.2 Gen 2, 2x USB 4.0, 2x USB 2.0, and a high-speed SD 4.0 card slot - making high-resolution image and video transfer seamless and efficient.

AMD is preparing the next-generation mobile processor based on Zen 6 "Medusa Point" APUs. According to industry leaker HXL, the new lineup will feature a surprisingly high core count, especially on the top-end Ryzen 9 model, which is expected to pack 22 CPU cores. Under the hood, the Ryzen 5 and Ryzen 7 Medusa Point APUs will use a hybrid layout that combines four high-performance classic cores, four dense cores optimized for efficiency, and two ultra-low‑power cores. On the graphics side, these chips will offer eight RDNA 3.5+ compute units, an upgrade from previous designs but still modest compared to some rivals. What really stands out is the Ryzen 9 variant, which reportedly adds a second Core Complex Die from AMD's desktop lineup. That extra 12‑core CCD joins the main ten-core die, bringing the total to 22 cores. While this pushes the CPU core count, AMD keeps the integrated graphics configuration unchanged, with just eight compute units.

Initial performance data hints that this trimmed-down GPU might run about 20 to 25 percent slower than the 12‑CU Radeon 890M used in earlier models. The eight CU GPU is half the power of current generation "Strix Point" with 16 CUs, but Medusa Point packs some RDNA 3.5+ upgrades. Even then, the raw CPU power should appeal to users who rely on heavily threaded applications, particularly in ultraportable laptops and mobile workstations. Medusa Point is built on the FP10 package, allowing for a slightly larger die size that houses the new core arrangement without sacrificing efficiency. Including dedicated low-power cores also marks AMD's first embrace of a heterogeneous x86 approach, similar to Intel's recent Core Ultra series. While more serious graphics work will mandate a combination of discrete GPUs for demanding workloads, Medusa Point will set new expectations for integrated CPU performance in thin, light notebooks.

Last August, the rumor mill indicated a possible arrival of Radeon RX 7400 and RX 7300 graphics cards—potentially distributed through OEM channels. Going further back—during the early days of RDNA 3—insiders posited that true entry-level models were on the way; sitting below the already lowly Radeon RX 7600 model in Team Red's "Navi 33" GPU-based stack. Around mid-2023, Kepler_L2 envisioned humble "Navi 33 Ultra Lite with 4 CUs" designs. Currently, TechPowerUp's database alludes to a "never released" Radeon RX 7500 XT card, but AMD's (presumably less potent) RX 7400 and RX 7300 options were not deemed worthy of similar treatment.

To the surprise of many industry watchdogs, quite fresh AIDA64 Extreme (beta) release notes—version: 7.65.7415 beta (May 05, 2025)—mention "AMD Radeon RX 7300 (Navi 33)" GPU information. VideoCardz highlighted this discovery with a quick investigative piece; they did not find any official listings of this mysterious identifier. Two other RDNA 3 era models were present in the latest AIDA64 update: an already launched "Radeon Pro W7700 (Navi 32)" card and 2024's "Radeon RX 7800M (Navi 32)" mobile solution. Despite the upcoming launch of lower end RDNA 4 graphics cards, Team Red and a couple of its board partners are still pushing out "new" Navi 33 GPU-based cards—most notably Chinese market exclusive Radeon RX 7650 GRE custom models. In theory, a large surplus of "Hotpink Bonefish" GPU dies would require depletion in 2025.

Kuroutoshikou—a Japanese PC hardware brand—has introduced a special Blade and Soul NEO collaboration Radeon RX 7600 8 GB model. The company's press release goes into great detail about their retail packaging being "decorated with the beautiful characters and gorgeous stages of Blade," but the card itself is not adorned with any fancy illustrations or markings. In fact, the bare black design seems to be borrowed almost directly from PowerColor's Fighter stable (minus two stickers). As pointed out by VideoCardz, Kuroutoshikou already offers a "normal edition" Radeon RX 7600 SKU—minus NCSoft-related material. An exclusive in-game bonus seems to be the main draw here—as disclosed in the company's marketing spiel: "if you purchase this product and apply on the special page, you will receive a campaign code for the limited costume 'Gentleness' that can be used in Blade and Soul NEO."

Kuroutoshikou will deliver their "RD-RX7600-8G/Blade & Soul NEO Collaboration" edition to market on April 4, which could place it in a similar release timeframe with natural successors—AMD and board partners are expected to launch Radeon RX 9060 Series graphics cards in the second quarter of 2025. Surprisingly, the lower end of RDNA 3 keeps generating news articles—PowerColor seems to be working on a "V2" Fighter Radeon RX 7600 SKU, while plenty of China market-exclusive Radeon RX 7650 GRE cards have appeared within the past month or two. A possible surplus of Navi 33 GPUs could explain the sudden "reemergence" of previous-gen hardware. Going back to Kuroutoshikou; their current graphics card portfolio seems to consist of slightly rebranded PowerColor designs for the AMD side of things, while NVIDIA GeForce offerings are rebadged GALAX IPs.

AMD's latest flagship APU, the Ryzen AI MAX+ 395 "Strix Halo," demonstrates some impressive performance advantages over Intel's "Lunar Lake" processors in large language model (LLM) inference workloads, according to recent benchmarks on AMD's blog. Featuring 16 Zen 5 CPU cores, 40 RDNA 3.5 compute units, and over 50 AI TOPS via its XDNA 2 NPU, the processor achieves up to 12.2x faster response times than Intel's Core Ultra 258V in specific LLM scenarios. Notably, Intel's Lunar Lake has four E-cores and four P-cores, which in total is half of the Ryzen AI MAX+ 395 CPU core count, but the performance difference is much more pronounced than the 2x core gap. The performance delta becomes even more notable with model complexity, particularly with 14-billion parameter models approaching the limit of what standard 32 GB laptops can handle.

In LM Studio benchmarks using an ASUS ROG Flow Z13 with 64 GB unified memory, the integrated Radeon 8060S GPU delivered 2.2x higher token throughput than Intel's Arc 140V across various model architectures. Time-to-first-token metrics revealed a 4x advantage in smaller models like Llama 3.2 3B Instruct, expanding to 9.1x with 7-8B parameter models such as DeepSeek R1 Distill variants. AMD's architecture particularly excels in multimodal vision tasks, where the Ryzen AI MAX+ 395 processed complex visual inputs up to 7x faster in IBM Granite Vision 3.2 3B and 6x faster in Google Gemma 3 12B compared to Intel's offering. The platform's support for AMD Variable Graphics Memory allows allocating up to 96 GB as VRAM from systems equipped with 128 GB unified memory, enabling the deployment of state-of-the-art models like Google Gemma 3 27B Vision. The processor's performance advantages extend to practical AI applications, including medical image analysis and coding assistance via higher-precision 6-bit quantization in the DeepSeek R1 Distill Qwen 32B model.

Throughout the early months of 2025, PowerColor's new product strategy seemed to signal the retirement of their entry-level "Fighter" graphics card series. A natural replacement—dubbed "Reaper"—was debuted in an official capacity at CES 2025, with AMD's introduction of the new-gen Radeon RX 9070 Series. With the delay of Team Red's RDNA 4 global market release to March 6, the Taiwanese manufacturer proceeded with a launch of custom Radeon RX 7650 GRE Reaper models in China around late February. According to a past weekend VideoCardz news report, PowerColor is expanding its RDNA 3 portfolio once again. Their investigation has unearthed a new Radeon RX 7600 "Fighter V2" model—confusingly, this variant seems to borrow the recently introduced dual-fan Reaper cooling solution.

According to leaked information, PowerColor is expected to launch its "V2" Fighter model to a global buying audience. The AIB's Radeon RX 7650 GRE Reaper cards (in black or white) will continue to serve as Chinese market exclusives. The leaked "RX7600 8G-F/V2" product identifier indicates that PowerColor is keeping its "Fighter" family alive for a little bit longer, perhaps with a designation of cards that utilize older generation architecture. Insiders propose that the incoming PowerColor RX 7600 Fighter V2 model is configured with reference specifications; mirroring version 1.0's credentials. VideoCardz has mocked up speculative packaging (see below); they reckon that a black variant is lined up for an imminent release—the aforementioned product code has cropped up across retail databases.

MSI has officially confirmed that it will not manufacture graphics cards based on AMD's latest RDNA 4 architecture, effectively bypassing the entire Radeon RX 9000 series lineup. In a statement to Tom's Hardware, an MSI representative briefly noted the company "is not manufacturing AMD GPUs this generation," pausing its AMD partnership while leaving the door open for future collaborations. Data compiled by Tom's Hardware shows MSI produced 45 distinct models during the RDNA 2 generation (RX 6000-series), but dramatically scaled back to just four custom designs for RDNA 3 (RX 7000-series)—representing a 91% reduction in AMD SKU diversity. Those limited RDNA 3 offerings, including the flagship RX 7900 XTX, notably reused cooling solutions from previous-generation AMD cards, indicating reduced R&D allocation compared to the company's NVIDIA lineup.

MSI's withdrawal from offering AMD-based solutions is due to several factors. NVIDIA's dominance in market share (83%, according to Steam hardware surveys) provides partners with stronger return-on-investment potential, while EVGA's 2022 exit from the GPU market created an opportunity for MSI to strengthen its position as a premier NVIDIA partner. There were reportedly delays in finalizing the MSRP for the RX 9000 series, which complicated manufacturers' calculations, production planning, and shipments. The vacuum left by MSI creates potential opportunities for other emerging players like Acer, which has been slowly expanding its AMD graphics card presence. However, MSI's decision—likely finalized during earlier strategic planning cycles—could potentially push away AMD enthusiasts despite strong projected demand for the Radeon RX 9070 series. Whether this represents a permanent strategic realignment or a temporary market response, we are yet to find out. AMD plans to return to the high-end GPU segment with its UDNA generation, so we have to wait and see if MSI makes a comeback here. For AIBs, partnerships with GPU makers are essential to get R&D resources behind a new product. Once partnerships pause, it is hard to get that wheel going again.

This month, we at AMD celebrate two significant milestones in the Radeon story. First, the 25th anniversary of Radeon, a journey that began in 2000 with the ATI Radeon DDR card. Back then, 32 MB of VRAM, a 143 MHz clocks, and 30M transistors were cutting-edge tools that sparked your early adventures. Today, those specs are a nostalgic memory, dwarfed by the leaps we've made together culminating in the 24 GB of memory, multi-GHz clocks, and nearly 60B transistors of RDNA 3 cards driving the immersive worlds you now explore. But we're not stopping there. We're proud to continue that innovation journey with the RDNA 4-based Radeon RX 9070 XT and Radeon RX 9070, available starting today. This is more than a new chapter for us, it's a promise to you, the gamers who fuel our passion. We know what matters when you choose your next GPU: raw performance to conquer your favorite titles, tech that's ready for tomorrow's blockbusters, and value that respects your investment. That's precisely what RDNA 4 delivers.

Our goal with RDNA 4 wasn't to chase an elite crown few can reach. Instead, we focused on you, the heart of gaming, crafting cards that bring exceptional power to the setups most of you run. Compared to our last gen, RDNA 4 boosts raster performance for crisper, smoother visuals. Ray tracing throughput doubles, letting you soak in lifelike lighting and reflections without compromise. And with an 8x uplift in machine learning performance, we're unlocking new possibilities - like FSR 4, our latest leap in ML-based upscaling.

PowerColor China has published two new Radeon RX 7650 GRE product pages on its website; indicating that more Reaper series graphics cards are on the way. TechPowerUp inspected an example at CES 2025, but that particular triple-fan model housed upcoming RDNA 4 GPU tech. PowerColor's brand-new "Reaper" family is rolling out in the near future; acting as a replacement to their familiar budget-friendly "Fighter" tier. The manufacturer released its Radeon RX 7650 GRE Hellhound (mid-range) card earlier this month—the standard black and Spectral White Reaper models are expected to arrive with slightly lower pricing, likely conforming to AMD's official MSRP guideline (2099 RMB/~$289 USD).

The latest report—from VideoCardz—suggests that PowerColor will be expanding its portfolio of Spectral White graphics card designs. Company insiders hinted about incoming pale Radeon RX 9070 XT options; namely a new pale Hellhound offering. The Taiwanese firm is reportedly considering the release of a Spectral White spin-off of their Radeon 9070 XT Red Devil model—potentially, the first of its kind to get a pale aesthetic treatment. PowerColor's demon-themed flagship card series—traditionally—sports black shroud and backplate designs, with red accents and lighting zones. A speculative designation of "Red Devil Spectral White" seems to be quite clumsy; so a snappier title could be in the works.

An alleged AMD Radeon RX 9070 XT graphics card has posted an overall score of 7931 points in a Furmark v2.5 OpenGL test session. Earlier today, Tomasz Gawroński shared a hastily prepared screenshot, accompanied by his observations: "I found an AMD (RDNA 4) Radeon RX 9070 XT GPU and Ryzen 9 9950X3D CPU with hacked drivers. Device ID (1002-7550) matches the recently leaked Geekbench entry. There are multiple benches with 9950X3D on Furmark. Scores are impressive: 41-48% higher than Radeon 7800 XT." VideoCardz believes that the Furmark leak points to the true potential of Team Red's upcoming Navi 48-based graphics cards. Recent Geekbench results—reportedly produced by Radeon RX 9070 XT and 9070 (non-XT) pre-release samples—have indicated underwhelming performance; closer to previous-gen mid-range levels.

The "hacked" Radeon RX 9070 XT sample's Furmark tally—of 7931—places it higher than previously perceived; when compared to Team Red's middle-to-high range portfolio of RDNA 3 offerings. VideoCardz posited that the leaked candidate's score: "puts it almost at the Radeon RX 7900 XTX's level, faster than the Radeon RX 7900 XT, RX 7900 GRE, and over 50% higher than the 7800 XT. Based on rumors we heard this week, AMD is said to be claiming over ~40% higher performance at 4K (games) than the 7900 GRE, so this would be in line with these claims."

Last week, AMD China and involved board partners launched region-exclusive Radeon RX 7650 GRE graphics card models; initial publicity produced a mountain of promotional imagery, but only a minority of AIBs published GPU specification sheets. On Tuesday (February 11), new Team Red documentation revealed a slightly confusing change in nomenclature—as of 2025, GRE stands for "Great Radeon Edition," rather than "Golden Rabbit Edition." On launch day, ASRock China showcased their Navi 33 GPU-based Radeon RX 7650 GRE Challenger 8 GB OC Edition card. A product page has appeared on the company's local website, complete with technical information. We are looking at 8 GB of GDDR6 VRAM, 2048 stream processors, a game clock frequency of 2539 MHz, and a boost clock of "up to 2810 MHz." The ASRock official store price is listed as 2049 RMB (~$281 USD).

Another model, the ASRock Radeon RX 7650 GRE Challenger OC+ Edition has appeared online; its official product page looks almost identical to the OC (non-plus) Edition's. The tacked-on "+" denotes key differences in game clock and boost clock factory settings—ASRock has implemented higher frequencies: 2400 MHz and 2725 MHz (respectively). After analyzing promotional imagery, the slightly fancier card appears to sport a chunkier shroud design. The new "Great Radeon Edition" OC+ Edition is not listed on ASRock's official e-tail outlet. Amusingly, the two product pages contain the same error; their "main specification" sections point to a non-existent AMD Radeon RX 7600 GRE GPU. Team Red's Radeon RX 7650 GRE GPU seems to be a slightly tweaked variant of the Radeon RX 7600 (Navi 33 XL) 8 GB model.

Mid-week, a member of Valve's Linux graphics driver team updated the Mesa 25.0 documentation with an insightful note—in which, Samuel Pitoiset (lead developer of the RADV open-source driver) shared an early observation regarding AMD's upcoming RDNA 4 generation. The software engineer and his colleagues are busy getting everything finalized prior to an impending Mesa 25.0 feature freeze. Wednesday's patch notes reveal the outlook for "GFX12 on RADV"—Pitoiset seemed satisfied about "initial support expectations—he reckoned that it: "should be good enough, but it's missing two features (cooperative matrix and video decode/encode), compared to GFX11 (RDNA 3) because lack of time. DCC is still under active development, but it might be possible to finish it during the RC period."

According to Phoronix's expert opinion, the two missing features are not complete "show-stoppers" for potential buyers of first wave RDNA 4 GPUs. The site's editor-in-chief—Michael Larabel—posits that the "vast majority of those wanting to buy the Radeon RX 9070 graphics cards (when they launch in March)" will not be discouraged by the inceptive absence of RADV Vulkan Video and VK_KHR_cooperative_matrix. He added some post-publication clarification regarding the RADV patch notes: "this is only about Vulkan Video, not VA-API video acceleration... It seems some readers are taking this to mean VA-API support for the new VCN block isn't ready for RDNA 4. It's just the RADV Vulkan Video support that isn't complete." Mesa 25.0 is expected to reach a stable release stage by the end of February—just ahead of Team Red's next-gen desktop GPU launch. Late last week, an AMD official divulged that their team would be: "taking a little extra time to optimize the software stack for maximum performance" on Radeon RX 9070 XT and Radeon RX 9070 (non-XT) graphics cards.

AMD today put out detailed guides on how to get DeepSeek R1 distilled reasoning models to run on Radeon RX graphics cards and Ryzen AI processors. The guide confirms that the new Ryzen AI Max "Strix Halo" processors come in hardwired to LPCAMM2 memory configurations of 32 GB, 64 GB, and 128 GB, and there won't be a 16 GB memory option for notebook manufacturers to cheap out with. The guide goes on to explain that "Strix Halo" will be able to locally accelerate DeepSeek-R1-Distill-Llama with 70 billion parameters on the 64 GB and 128 GB memory configurations of "Strix Halo" powered notebooks, while the 32 GB model should be able to run DeepSeek-R1-Distill-Qwen-32B. Ryzen AI "Strix Point" mobile processors should be capable of running DeepSeek-R1-Distill-Qwen-14B and DeepSeek-R1-Distill-Llama-14B on their RDNA 3.5 iGPUs and NPUs. Meanwhile, older generation processors based on "Phoenix Point" and "Hawk Point" chips should be capable of DeepSeek-R1-Distill-Llama-14B. The company recommends running all of the above distills in Q4 K M quantization.

Switching gears to the discrete graphics cards, and AMD is only recommending its Radeon RX 7000 series for now, since the RDNA 3 graphics architecture introduces AI accelerators. The flagship Radeon RX 7900 XTX is recommended for DeepSeek-R1-Distill-Qwen-32B distill, while all SKUs with 12 GB to 20 GB of memory—that's RX 7600 XT, RX 7700 XT, RX 7800 XT, RX 7900 GRE, and RX 7900 XT, are recommended till DeepSeek-R1-Distill-Qwen-14B. The mainstream RX 7600 with its 8 GB memory is only recommended till DeepSeek-R1-Distill-Llama-8B. You will need LM Studio 0.3.8 or later and Radeon Software Adrenalin 25.1.1 beta or later drivers. AMD put out first party LMStudio 0.3.8 tokens/second performance numbers for the RX 7900 XTX, comparing it with the NVIDIA GeForce RTX 4080 SUPER and the RTX 4090.

Jon Peddie Research (JPR) reports that shipments of integrated and standalone GPUs exceeded 251 million units in 2024, marking a 6% year-over-year rise. This growth continues to outpace the shipment figures for client CPUs, given that most desktop and laptop processors include built-in graphics, and that leading suppliers such as AMD and NVIDIA also provide tens of millions of discrete GPUs annually. Thanks to some earlier per-quarter data, we find that in the desktop segment, JPR shows that manufacturers shipped about 18.2 million discrete graphics cards in the first half of 2024—a 46% increase compared to the same period in 2023. However, shipments slowed in the third quarter, falling to 8.1 million units, down from 8.9 million units in Q3 2023.

Analysts partly attribute this dip to inventory adjustments at AMD, along with the end of product cycles for Ada Lovelace and RDNA 3 GPUs. Typically, the fourth quarter sees a boost in discrete GPU purchases as gamers gear up for new software releases. Yet it is not entirely clear whether that trend held through late 2024. Early indications suggest that total graphics card shipments for the year lightly exceed the totals from 2023, thanks primarily to strong demand in the first half. We still need data from Q4 of 2024 to see the trend in its final months of the year, despite knowing that the year-over-year growth is 6%. Nonetheless, current levels appear unlikely to approach the elevated peaks observed in 2021 or 2022.

We had a chance to get close and personal with the new ZOTAC ZONE GAMING prototype at the CES 2025 show. While it is not a final product, we had a chance to see it in action as it is a working prototype and thanks to updated hardware, it should provide much higher performance compared to the ZOTAC Zone that we had a chance to review last year.

The biggest update is the 4 nm Ryzen AI 9 HX 370 processor. The Strix Point architecture brings 12-core/24-thread CPU (four performance and eight efficiency) based on Zen 5 architecture, 24 MB of shared L3 cache and 1 MB of L2 cache per core. It also comes with Radeon 890M, a RDNA 3.5 architecture GPU with 16 Compute Units. ZOTAC also increased the amount of LPDDR5X RAM to 32 GB and raised M.2 2280 PCIe 4.0 SSD storage space to 1 TB. The Ryzen AI 9 HX 370 has a standard TDP of 28 W, and a configurable TDP between 15 W and 54 W, so it gives ZOTAC a lot of room to work with. It also features 50 TOPS XDNA NPU.

AMD today vastly fleshed out its mobile processor lineup with the introduction of two new processor lines besides the Ryzen AI Max 300 series. This includes the introduction of more processor models in the Ryzen AI 300 series that are powered by the "Strix Point" silicon, and the introduction of the Ryzen 200 series mobile processors, which are based on the older "Hawk Point" silicon. In 2024, AMD had debuted the Ryzen AI 300 series "Strix Point," but with just the top-end Ryzen AI 9 370 and 365, which came with maxed out 12-core/24-thread (4x Zen 5 + 8x Zen 5c) core configuration, and a maxed out iGPU with 16 CU. Today the company is introducing the Ryzen AI 7 350, the Ryzen AI 5 340, and their AMD PRO variants for commercial notebooks. Both the consumer and commercial parts have identical specs, except for the latter featuring the AMD PRO feature-set.

The Ryzen AI 7 350 comes with a CPU configuration of 8-core/16-thread (4x Zen 5 + 4x Zen 5c). All cores have a base frequency of 2.00 GHz, the Zen 5 cores boost up to 5.00 GHz. The iGPU on offer is the Radeon 860M, with 12 CU and an engine clock of up to 3.00 GHz. TDP is configurable between 15 W to 55 W. The Ryzen AI 5 340 comes with a 6-core/12-thread configuration (3x Zen 5 + 3x Zen 5c), and CPU clock speeds of 2.00 GHz base with 4.80 GHz boost achievable on the Zen 5 cores. The iGPU is heavily cut down, with just 4 CU available, and an iGPU engine clock of 2.90 GHz. Notebook designers can configure this chip with a wide power range from 15 W to 55 W. All four processor models mentioned above come with a Ryzen AI XDNA 2 NPU that's capable of 50 AI TOPS, which means they're all Microsoft Copilot+ AI PC logo eligible.

According to Tweakers, AMD's Radeon RX 7900 GRE graphics card has reached end-of-life status, as confirmed by multiple AMD board partners they have contacted. The announcement comes just months after the card's expansion into European markets following its initial 2023 exclusive launch in China. Tweakers report that the supply of the RX 7900 GRE is rapidly declining across retail channels. While ASUS models remain somewhat available, the manufacturer has informed Tweakers that deliveries are currently "limited." AMD has not responded to their multiple requests for comment regarding the discontinuation. The RX 7900 GRE offers compelling specifications that position it as a slightly scaled-down variant of the more premium RX 7900 XT.

Built on AMD's RDNA 3 architecture, the card features 80 CUs and 16 GB of GDDR6 memory and operates at a 260 W TDP. The timing of this discontinuation is particularly interesting as AMD prepares to unveil its next-generation RDNA 4-based Radeon RX 8000 series. Perhaps AMD is trying to flush out its remaining inventory to make room for its Radeon RX 8000 series GPUs, which should mainly target the middle-range of the next-generation GPU families, including competition like NVIDIA with "Blackwell" and Intel with "Battlemage." With the new card scheduled to appear during AMD's CES keynote on January 6 in Las Vegas, we have to wait and see what products AMD puts out before analyzing why AMD decided to EOL its Radeon RX 7900 GRE.

While the current generation of AMD graphics cards employs RDNA 3 at its core, and the upcoming RX 8000 series will feature RDNA 4, the latest leaks suggest RDNA 5 is not in development. Instead, UDNA will succeed RDNA 4, simplifying AMD's GPU roadmap. A credible source on the Chiphell forums, zhangzhonghao, reports that the UDNA-based RX 9000 series and Instinct MI400 AI accelerator will incorporate the same advanced Arithmetic Logic Unit (ALU) designs in both products, reminiscent of AMD's earlier GCN architectures before the CDNA and RDNA split. Sony's next-generation PlayStation 6 is also rumored to adopt UDNA technology. The PS5 and PS5 Pro currently utilize RDNA 2, while the Pro variant integrates elements of RDNA 4 for enhanced ray tracing. The PS6's CPU configuration remains unclear, but speculation revolves around Zen 4 or Zen 5 architectures.

The first UDNA gaming GPUs are expected to enter production by Q2 2026. Interestingly, AMD's RDNA 4 GPUs are anticipated to focus on entry-level to mid-range markets, potentially leaving high-end offerings until the UDNA generation. This strategic pause may allow AMD to refine AI-accelerated technologies like FidelityFX Super Resolution (FSR) 4, aiming to compete with NVIDIA's DLSS. This unification is inspired by NVIDIA's CUDA ecosystem, which supports cross-platform compatibility from laptops to high-performance servers. As AMD sees it, the decision addresses the challenges posed by maintaining separate architectures, which complicate memory subsystem optimizations and hinder forward and backward compatibility. Putting developer resources into RDNA 5 is not economically or strategically wise, given that UDNA is about to take over. Additionally, the company is enabling ROCm software support across all products ranging from consumer Radeon to enterprise Instinct MI. Accelerating software for one platform will translate to the entire product stack.

GIGABYTE TECHNOLOGY Co. Ltd, a leading manufacturer of premium gaming hardware, today launched the cutting-edge GIGABYTE AMD Radeon PRO W7800 AI TOP 48G. GIGABYTE has taken a significant leap forward with the release of the Radeon PRO W7800 AI TOP 48G graphics card, featuring AMD's RDNA 3 architecture and a massive 48 GB of GDDR6 memory. This significant increase in memory capacity, compared to its predecessor, provides workstation professionals, creators, and AI developers with incredible computational power to effortlessly handle complex design, rendering, and AI model training tasks.

⁠GIGABYTE stands as the AMD professional graphics partner in the market, with a proven ability to design and manufacture the entire Radeon PRO series. Our dedication to quality products, unwavering business commitment, and comprehensive customer service empower us to deliver professional-grade GPU solutions, expanding user's choices in workstation and AI computing.⁠

AMD has published a blog post about its latest AMD Ryzen AI 300 series processors, claiming they are changing the game for portable devices. To back these claims, Team Red has compared its Ryzen AI 9 HX 370 processor to Intel's latest Core Ultra 7 258V, using the following games: Assassin's Creed Mirage, Baldur's Gate 3, Borderlands 3, Call of Duty: Black Ops 6, Cyberpunk 2077, Doom Eternal, Dying Light 2 Stay Human, F1 24, Far Cry 6, Forza Horizon 5, Ghost of Tsushima, Hitman 3, Hogwarts Legacy, Shadow of the Tomb Raider, Spider-Man Remastered, and Tiny Tina's Wonderlands. The conclusion was that AMD's Ryzen AI 9 HX 370, with its integrated Radeon 890M graphics powerhouse, outperformed the Intel "Lunar Lake" Core Ultra 7 258V with Intel Arc Graphics 140V by 75% on average.

To support this performance leap, AMD also relies on software technologies, including FidelityFX Super Resolution 3 (FSR 3) and HYPR-RX, to unlock additional power and gaming efficiency. FSR 3 alone enhances visuals in over 95 games, while HYPR-RX, with features like AMD Fluid Motion Frames 2 (AFMF 2) and Radeon Anti-Lag, provides substantial performance boosts across thousands of games. The company has also compared its FSR/HYPR-RS combination with Intel's XeSS, which is available in around 130 games. AMD claims its broader suite supports 415+ games and is optimized for smoother gameplay. The AFMF 2 claims support with thousands of titles, while Intel's GPU software stack lacks a comparison point. Of course, these marketing claims are to be taken with a grain of salt, so independent testing is always the best to compare the two.

ASUS over the weekend introduced its third custom-design AMD Radeon RX 7600 graphics card, the RX 7600 DUAL EVO OC. This card is visually the largest, most spruced-up custom-design for this GPU from the company's DUAL series. It's positioned a notch below the premium ROG Strix RX 7600 OC, which remains the only ROG branded graphics card based on the RX 7000 series. The DUAL EVO OC is 22.9 cm in length, and 12.3 cm in height, and is 2.5 slots thick. Its cooling solution features an aluminium fin-stack heatsink, which is ventilated by a pair of 80 mm Axial-Tech fans. The cooler offers idle fan-stop.

Out of the box, the ASUS RX 7600 DUAL EVO OC comes with a 2280 MHz Game clock, and 2695 MHz maximum boost frequency, compared to AMD reference clock speeds of 2250 MHz Game clocks and 2655 MHz boost. The card draws power from a single 8-pin PCIe power connector, and puts out display outputs that include three DisplayPort 1.4a, and one HDMI 2.1. Based on the 6 nm "Navi 33" silicon and driven by the RDNA 3 graphics architecture, the RX 7600 features 2,048 stream processors across 32 CU, along with 64 AI accelerators, and 32 ray accelerators. It comes with 8 GB of 18 Gbps GDDR6 memory across a 128-bit wide memory bus. ASUS didn't reveal pricing.

Our latest front-page poll sheds light on what people want from the next generation of gaming GPUs. We asked our readers what mattered most to them, with answers including raster performance, ray tracing performance, energy efficiency, upscaling or frame-gen technologies, the size of video memory, and lastly, pricing. Our poll ran from September 19, and gathered close to 24,000 votes as of this writing. Pricing remains the king of our polls, with the option gathering 36.1% of the vote, or 8,620 votes. Our readers expect pricing of next-generation GPUs to remain flat, variant-for-variant, and not continue on the absurdly upward trend it has had for the past few generations, with the high-end being pushed beyond the $1,000-mark, and $500 barely bringing in a 1440p-class GPU, while 4K-capable game consoles exist.

Both AMD and NVIDIA know that Moore's Law is cooked, and that generational leaps in performance and transistor counts are only possible with increase in pricing for the latest foundry nodes. AMD even tried experimenting with disaggregated (chiplet-based) GPUs with its latest RDNA 3 generation, before calling it quits on the enthusiast-segment, so it could focus on the sub-$1000 performance segment. The second most popular response was Raster 3D performance (classic 3D rendering performance), which scored 27% or 6,453 votes.

AMD is readying a new mainstream graphics card positioned based on its current RDNA 3 graphics architecture, the Radeon RX 7650 GRE (Golden Rabbit Edition). The company has had great success in selling graphics card SKUs with the "GRE" brand extension in China, with the RX 6750 GRE being a popular SKU there; and it even has an enthusiast-class SKU with the RX 7900 GRE, which saw a global launch earlier this year. The company is hoping for the "GRE" moniker to compete better against the GeForce RTX 4060, at least in specific markets. A new Benchlife.info report says that the RX 7650 GRE will be based on the 6 nm "Navi 33" monolithic silicon, and not the 5 nm "Navi 32" chiplet-based GPU previously reported.

AMD has already maxed out the "Navi 33" for both the RX 7600 and RX 7600 XT, with the latter only seeing its memory size doubled over the former, so it remains to be seen where AMD goes with the RX 7650 GRE. The "7650" numbering suggests a faster SKU, so it's possible that AMD increases the engine clock speeds of the "Navi 33" by as much as it can. The RX 7600 comes with a 2.25 GHz Game clock, which the RX 7600 XT slightly bumps up to 2.47 GHz. If we were to guess, the RX 7650 GRE could focus on increasing the Game clock, not the memory size; and so it could have the power configuration of the RX 7600 XT, and room for Game clocks either on-par or higher than the RX 7600 XT, while retaining the 8 GB memory size of the RX 7600.