TechPowerUP NEWS

AMD and Stability AI made an announcement today about their joint effort to adapt Stable Diffusion 3.0 Medium for Stability Amuse AI art creator. This move allows for better picture generation and text handling abilities to run on Neural Processing Units (NPUs). Earlier, at Computex 2024, AMD showed off the first-ever block FP16 stable diffusion model in the world, SDXL Turbo, which they created with Stability AI. This new model reached FP16 accuracy at INT8 performance levels. The two companies now introduce their newest teamwork featuring the first block FP16 SD 3.0 Medium model. This step forward brings higher picture quality while being tailored for AMD Ryzen AI XDNA 2 NPUs. People can use this right away through Amuse 3.1 by Tensorstack running in HQ mode on Ryzen AI-powered laptops that support it.

The new block FP16 SD 3 Medium model needs less memory than older versions as it works on 24 GB laptops while using up 9 GB of memory. This data type enables high-quality local AI image generation on laptops with less than 32 GB of memory using high precision format without excessive quantization. The system incorporates a two-stage pipeline powered by the AMD XDNA 2 NPU that enhances the standard 2 megapixel (1024 x 1024) output images to 4 megapixel (2048 x 2048) resolution. In the past, Amuse offered Stable Diffusion Medium, but it needed GPU processing, which held back widespread use. Now, users can pick between GPU or NPU processing. To test this new feature, users need to download and install the latest Adrenalin Driver, then the Amuse 3.1 Beta, and while in EZ Mode, move the slider to HQ and toggle XDNA 2 Stable Diffusion Offload.[Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

PLAION REPLAI, a leading global publisher and developer of retro video game software and hardware, and Atari, one of the world's most iconic consumer brands and interactive entertainment producers, are proud to announce today that the Atari 2600+ PAC-MAN Edition - a newly developed version of the classic console - will be launching worldwide on 31st October 2025. This exciting new product also celebrates the long-standing collaboration between Atari and Bandai Namco Entertainment America Inc., with Namco being the original developer of PAC-MAN.

In tribute to the original video game romance that brought PAC-MAN to home consoles for the first time on the Atari 2600, the new hardware is infused with PAC-MAN's signature style. To compliment the release, a range of colorful joysticks will also be separately available.[Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...]
[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

For a limited time, use code FAMPLAN to [Only registered and activated users can see links. Click Here To Register...]—block pop-ups and keep your family safe, without monthly fees. Ad blockers have evolved from a nice-to-have to a must-have in today's internet landscape. With constant pop-ups, privacy threats, and page-clogging ads, tools like AdGuard can help make browsing smoother and safer, and you can protect up to nine devices for life. Right now, the lifetime AdGuard Family Plan is just $15.97 (MSRP $39.97) with code FAMPLAN at checkout.

Why is AdGuard Family Plan Worth It?

• Clean up your screens: Removes all types of ads, including banners, pop-ups, and video ads, such as those from YouTube, in most browsers.
• Keep your data private: Blocks trackers and activity analyzers that can follow you across websites.
• Stay safer online: Filters out known phishing sites and malware threats.

[Only registered and activated users can see links. Click Here To Register...]
[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

The 96-core/192-thread AMD Ryzen Threadripper PRO 9995WX was made to run at 5.00 GHz all-core, which no small feat considering its enormous core-count. This was pulled off by American overclocker SkywalkerAMD. The processor set a huge Cinebench R23 multithreaded score of 186800 points, but here's where it gets interesting—the processor was measured to draw a whopping 948 W. Cooling the beast is an SP5-compatible all in one liquid CPU cooler, but when faced with a thermal load of nearly 1 kW, the chip nears its T-junction max value, measuring 98 °C. The 186k points score is interestingly not the highest ever score for a 1P machine, that distinction goes to a workstation powered by a previous-generation Threadripper PRO 7995WX cooled by liquid nitrogen. Find more details in the HWBOT link below.[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

ASUS over the weekend released its China-exclusive special edition of the ROG Astral GeForce RTX 5080 graphics card that's co-branded with the Anime character Hatsune Miku. This isn't just a sticker job, but a complete re-skin, with the various metal accents on the cooler shroud and backplate being color-matched with the scheme, including two of the four Axial-Tech fans coming in turquoise-blue. Like what you see? Get ready to foot an over 100% premium over the NVIDIA MSRP for the RTX 5080, with ASUS pricing the card at an eye-watering RMB ¥16,999 (USD $2,369), which is double not just over the NVIDIA MSRP, but also the lowest street prices for the GPU. The card still falls short of the street pricing of the GeForce RTX 5090D, which can be found well above ¥18,000.[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

During the RISC-V Summit in China, NVIDIA's Frans Sijstermans, vice president of hardware engineering, officially announced that NVIDIA has ported its massive CUDA software library to RISC-V ISA, signaling that the company sees RISC-V adoption accelerating in every aspect of the hardware stack, and that it could be a perfect match for NVIDIA GPUs. Historically, NVIDIA ported its CUDA platform to every major instruction set, including x86, Arm, PowerPC, and even Sun's SPARC. This is done to ensure that if any platform takes off in the enterprise or any segment, NVIDIA is there from the start to support it and allow its GPUs to run at full speed. Now, the [Only registered and activated users can see links. Click Here To Register...] is making a bet that RISC-V will be the next desirable ISA to bring full support for.

NVIDIA envisions systems where its GPUs are the center of any acceleration, and supplementary RISC‑V CPUs oversee CUDA drivers, application logic, and the operating system, orchestrating parallel workloads entirely within the CUDA ecosystem. The diagram shown at the summit highlights a DPU handling networking tasks, creating a cohesive trio of compute, control, and data‑movement elements. If you recall, NVIDIA already uses NV-RISC-V CPU inside its GPUs for handling control logic on the GPU itself. That demonstrates NVIDIA's strategy to build heterogeneous platforms that combine RISC-V controllers with its GPUs, DPUs, and networking silicon. Now that CUDA is fully supported on RISC-V, NVIDIA could look into alternatives for its Grace CPU to be built on the RISC-V ISA. As the open-source ISA slowly breaks into the server space with the RVA23 specification, which NVIDIA mandates for CUDA support, we could see some interesting heterogeneous designs.[Only registered and activated users can see links. Click Here To Register...]
[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

EA DICE is confirmed to be working on a new "Battlefield" game, and it is expected to be called "Battlefield 6," according to a teaser by Doom49 on X. Doom49 claims to have received a "Battlefield 6" hamper in the mail from EA. The game is hence confirmed to be in development, although nowhere near release, or even play-testing. It's being reported on Reddit that the game will bring back a single-player campaign, a key component found missing in "Battlefield 2042" that faced limited success in the market. There is no word on the general setting or artistic backdrop of BF6—whether it's based on a near-future conflict like in "2042," or past conflicts like in "Battlefield V" and "Battlefield 1," or whether the campaign goes back to what worked best—a contemporary conflict setting as seen with "Bad Company 2," BF3 and BF4.[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

AMD's AI-driven FSR 4 upscaler has been limited to the relatively small number of titles released alongside the RX 9000 series until now. A weekend patch to the open-source utility OptiScaler changes this by adding full redirection support for FSR 4 in version 2025.7.19. Any DirectX 11 or DirectX 12 game that originally shipped with FSR 2, DLSS 2, or Intel XeSS can now utilize AMD's latest libraries instead. Early adopters can take advantage of FSR 4 enhancements, provided they have an RDNA 4 GPU and are willing to follow a few manual steps. You simply drop OptiScaler's FSR 4 proxy DLL into the game's installation folder, open OptiScaler's interface or its configuration tool to select "fsr4" as the backend, and turn off any built‑in anti‑lag or frame‑generation module to prevent timing conflicts between two low‑latency systems.

Once these steps are complete and you are running on DirectX 11 or 12 without an incompatible anti‑cheat or a Vulkan render path, you can launch the game, and OptiScaler will seamlessly drive FSR 4 in place of the original upscaler. There are still some caveats. OptiScaler does not yet support Vulkan render paths, and online titles protected by Easy Anti‑Cheat or BattlEye will refuse to launch when they detect the proxy DLL. Additionally, there is no universal installer, so each game directory must be modified individually. Even so, for single‑player or LAN‑only releases, this tool immediately extends FSR 4's reach from the current 65 native titles to potentially hundreds of legacy games.[Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

With its next generation Xe3 "Celestial" graphics architecture, Intel is returning to entry-level variants of its graphics IP that lacks features such as ray tracing, Kepler_L2 found in open source driver code supporting the integrated graphics solution of the upcoming "Wildcat Lake" mobile processor. Intel is designing "Wildcat Lake" to be its new entry-level mobile processor targeting Windows notebooks and Chromebooks priced well below the $400 mark. The iGPU of this chip, while based on the next-generation Xe3 architecture, only comes with 2 Xe cores worth 32 EU (execution units), which doesn't result in any meaningful amount of shader power to back ray traced workloads.

It might be premature to call this a true "Xe3-LP" sub-architecture at this point. While the original Xe-LP architecture powering the iGPU of "Tiger Lake," "Rocket Lake," and "Alder Lake" lacks ray tracing, it also physically lacks ray tracing units, and XMX cores for advanced matrix math acceleration driving AI workloads. It's being suggested in Kepler_L2's thread that the iGPU physically has RT units and XMX cores, but these are disabled, and are essentially deadweights on the silicon.[Only registered and activated users can see links. Click Here To Register...]
[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

According to one of the most reliable AMD leakers, Kepler_L2, AMD's upcoming UDNA (or RDNA 5) GPU generation will reintroduce higher-end GPU configurations with up to 96 Compute Units (CUs) in the top-end Navi 5X SKU, paired with a 384-bit bus for memory. We still don't know what type of memory AMD will ultimately use, but an early assumption could be that GDDR7 is on the table. When it comes to the middle of the stack, AMD plans a GPU with 64 CUs and a 256-bit memory bus, along with various variations of CUs and memory capacities around that. For the entry-level models, AMD could deliver a 32 CU configuration paired with a 128-bit memory bus. So far, memory capacities are unknown and may be subject to change as AMD finalizes its GPU lineup.

After the RDNA 4 generation, which left AMD without a top-end contender, fighting for the middle-end market share, a UDNA / RDNA 5 will be a welcome addition. We are looking forward to seeing what UDNA design is capable of and what microarchitectural changes AMD has designed. Mass production of these GPUs is expected in Q2 2026, so availability is anticipated in the second half of 2026. A clearer picture of the exact memory capacity and type will emerge as we approach launch.[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

Colorful Technology Company Limited, a leading brand in gaming PC components, gaming laptops, and Hi-fi audio products, proudly introduces the new iGame Shadow II DDR5 memory series, engineered to push the limits of performance for the latest AMD Ryzen 9000 series platforms. Building upon the success of its predecessor, the iGame Shadow II lineup delivers higher frequencies, tighter timings, and enhanced overclocking potential, ensuring gamers and creators can maximize their system's capabilities.

The iGame Shadow II flagship DDR5 memory kits build on the successful foundation of the first-generation iGame Shadow series, engineered specifically for the AMD Ryzen 9000 platform. It supports both AMD EXPO and Intel XMP 3.0 technologies. Featuring DDR5-6400 with ultra-low CL28 latency, these kits are designed for effortless plug-and-play performance with pre-configured profiles that unlock the full gaming potential of AMD's latest desktop CPUs. In a move that highlights its extreme overclocking capabilities, the memory operates at a high-performance voltage of 1.4 V, even lower than the already efficient 1.45 V of similar high-speed kits. This configuration offers enthusiasts greater flexibility and stability when pushing frequency and timings to the limit.[Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...]
[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

Premier embedded computing provider AAEON, has announced the release of the PICO-TWL4, a compact Pico-ITX board powered by Intel Processor N-series processors (formerly Twin Lake). The board is available in SKUs featuring both the 6 W, quad-core Intel Processor N150 and the 15 W, octa-core Intel Core 3 Processor N355. On its ultra-small 100 mm x 72 mm PICO-ITX form factor, the PICO-TWL4 leverages the optimized single and multi-thread performance of the newest efficiency-focused Intel processing platform while retaining the robust industrial design that AAEON's Pico-ITX line is known for.

The board offers up to 16 GB of DDR5 memory via SODIMM slot, two LAN ports (2.5GbE and 1GbE), and six USB interfaces, made up of two physical USB 3.2 Gen 2 ports and four USB 2.0 via pin header. For display, the PICO-TWL4 has both LVDS and HDMI, with the board's LVDS co-layed with eDP, providing greater variety to users based on project requirements.[Only registered and activated users can see links. Click Here To Register...]
[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

Today, NVIDIA unveiled OpenReasoning-Nemotron, a quartet of distilled reasoning models with 1.5B, 7B, 14B, and 32B parameters, all derived from the 671B-parameter DeepSeek R1 0528. By compressing that massive teacher into four leaner Qwen‑2.5-based students, NVIDIA is making advanced reasoning experiments accessible even on standard gaming rigs, without the need to worry about hefty GPU bills and cloud usage. The key is not some elaborate trick but raw data. Using the NeMo Skills pipeline, NVIDIA generated five million math, science, and code solutions, and then fine-tuned each one purely with supervised learning. Already, the 32B model hits an 89.2 on AIME24 and 73.8 on the HMMT February contest, while even the 1.5B variant manages a solid 55.5 and 31.5.

NVIDIA envisions these models serving as a powerful research toolkit. All four checkpoints will be available for download on Hugging Face, providing a strong baseline for exploring reinforcement-learning-driven reasoning or customizing the models for specific tasks. With GenSelect mode (which takes multiple passes for each question), you can spawn multiple parallel generations and pick the best answer, pushing the 32B model to outstanding performance that rivals or even exceeds OpenAI's o3‑high performance on several math and coding benchmarks. Since NVIDIA trained these models with supervised fine-tuning only, without reinforcement learning, the community has clean, state-of-the-art starting points for future RL experiments. For gamers and at-home enthusiasts, we get a model that can be very close to the state-of-the-art, entirely locally, if you have a more powerful gaming GPU.[Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

After a couple of slow weeks, we have a crammed week of new releases ahead of us and it kicks off with an action RPG set in feudal China. This is followed by a kingdom builder, a space PvP shooter with a twist or two, a bicycle race, bicycle race, bicycle race—sorry—a FPS zombie shooter and landing in a Sci-Fi noir mystery. There are a few more new games that didn't quite make the list as well.

[Only registered and activated users can see links. Click Here To Register...] / This week's major release / Thursday 24 July
Become Wuchang, a skilled pirate warrior struck by amnesia, who must navigate the uncertainties of her mysterious past while afflicted by the horrific Feathering disease. Explore the depths of Shu, enhancing your arsenal and mastering new skills harvested from fallen foes. Evolve your fighting style by sacrificing elusive Red Mercury and further develop new techniques obtained throughout your action-packed journey. Augment the weapons in your repertoire with powerful enchantments, allowing for a unique approach to combat and strategy that suits your personal style. [Only registered and activated users can see links. Click Here To Register...][Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...]
[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

Linux enthusiasts gather to hear the latest news: Intel is officially shutting down its Clear Linux distribution after ten years of development and optimizations. As many recall, Clear Linux is an Intel-optimized Linux distribution that serves as a high-performance, optimized OS designed to extract every last ounce of performance from Intel hardware, especially Intel Xeons. As software optimizations, such as AVX-512, became more common, Intel consistently pushed these optimizations and specific pre-compiled software with compiler flags and kernel optimizations that increased performance on Intel and even AMD CPUs by a few percents. Traditional Linux distributions weren't optimized in that regard as they were mostly compiled to run on every x86 or Arm CPU, without any special flags raised for performance. This has enabled compatibility, but lacks a few percentage points of performance that Clear Linux managed to extract. As of now, these efforts are essentially terminated, and the distribution will cease to receive security and quality of life improvements, effective immediately. Users are advised to switch to another distribution. The complete statement follows.[Only registered and activated users can see links. Click Here To Register...]
[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]