Just a few months after hiring Ben Skeggs, a lead maintainer of the open-source NVIDIA GPU driver for Linux kernel, NVIDIA has announced a complete transition to open-source GPU kernel modules in its upcoming R560 driver release for Linux. This decision comes two years after the company's initial foray into open-source territory with the R515 driver in May 2022. The tech giant began focusing on data center compute GPUs, while GeForce and Workstation GPU support remained in the alpha stages. Now, after extensive development and optimization, NVIDIA reports that its open-source modules have achieved performance parity with, and in some cases surpassed, their closed-source counterparts. This transition brings a host of new capabilities, including heterogeneous memory management support, confidential computing features, and compatibility with NVIDIA's Grace platform's coherent memory architectures.

The move to open-source is expected to foster greater collaboration within the Linux ecosystem and potentially lead to faster bug fixes and feature improvements. However, not all GPUs will be compatible with the new open-source modules. While cutting-edge platforms like NVIDIA Grace Hopper and Blackwell will require open-source drivers, older GPUs from the Maxwell, Pascal, or Volta architectures must stick with proprietary drivers. NVIDIA has developed a detection helper script to guide driver selection for users who are unsure about compatibility. The shift also brings changes to NVIDIA's installation processes. The default driver version for most installation methods will now be the open-source variant. This affects package managers with the CUDA meta package, run file installations and even Windows Subsystem for Linux.

GIGABYTE TECHNOLOGY Co. Ltd, a leading manufacturer of motherboards, graphics cards, and hardware solutions, released the GIGABYTE exclusive groundbreaking AI TOP Utility. With reinvented workflows, user-friendly interface, and real-time progress monitoring, AI TOP Utility provides a reinventing touch of local AI model training and fine-tuning. It features a variety of groundbreaking technologies that can be easily adapted by beginners or experts, for most common open-source LLMs, in anyplace even on your desk.

GIGABYTE AI TOP is the all-round solution for local AI Model Fine-tuning. Running local AI training and fine-tuning on sensitive data can relatively provide greater privacy and security with maximum flexibility and real-time adjustment. Collocating with GIGABYTE AI TOP hardware and AI TOP Utility, the common constraints of GPU VRAM insufficiency when trying to execute AI fine-tuning locally can be addressed. By GIGABYTE AI TOP series motherboard, PSU, and SSD, as well as GIGABYTE graphics cards lineup covering NVIDIA GeForce RTX 40 Series, AMD Radeon RX 7900 Series, Radeon Pro W7900 and W7800 series, the size of open-source LLM fine-tuning can now reach up to 236B and more.

TrendForce's latest industry report on AI servers reveals that high demand for advanced AI servers from major CSPs and brand clients is expected to continue in 2024. Meanwhile, TSMC, SK hynix, Samsung, and Micron's gradual production expansion has significantly eased shortages in 2Q24. Consequently, the lead time for NVIDIA's flagship H100 solution has decreased from the previous 40-50 weeks to less than 16 weeks.

TrendForce estimates that AI server shipments in the second quarter will increase by nearly 20% QoQ, and has revised the annual shipment forecast up to 1.67 million units—marking a 41.5% YoY growth.

Intel is preparing to launch its next generation of mobile CPUs with Core Ultra 200 series "Lunar Lake" leading the charge. However, as these processors are about to hit the market, leakers reveal Intel's plans for the next-generation Core Ultra 300 series "Panther Lake". According to rumors, Panther Lake will double the core count of Lunar Lake, which capped out at eight cores. There are several configurations of Panther Lake in the making based on the different combinations of performance (P) "Cougar Cove," efficiency (E) "Skymont," and low power (LP) cores. First is the PTL-U with 4P+0E+4LP cores with four Xe3 "Celestial" GPU cores. This configuration is delivered within a 15 W envelope. Next, we have the PTL-H variant with 4P+8E+4LP cores for a total of 16 cores, with four Xe3 GPU cores, inside a 25 W package. Last but not least, Intel will also make PTL-P SKUs with 4P+8E+4LP cores, with 12 Xe3 cores, to create a potentially decent gaming chip with 25 W of power.

Intel's Panther Lake CPU architecture uses an innovative design approach, utilizing a multi-tile configuration. The processor incorporates five distinct tiles, with three playing active roles in its functionality. The central compute operations are handled by one "Die 4" tile with CPU and NPU, while "Die 1" is dedicated to platform control (PCD). Graphics processing is managed by "Die 5", leveraging Intel's Xe3 technology. Interestingly, two of the five tiles serve a primarily structural purpose. These passive elements are strategically placed to achieve a balanced, rectangular form factor for the chip. This design philosophy echoes a similar strategy employed in Intel's Lunar Lake processors. Panther Lake is poised to offer greater versatility compared to its Lunar Lake counterpart. It's expected to cater to a wider range of market segments and use cases. One notable advancement is the potential for increased memory capacity compared to Lunar Lake, which capped out at 32 GB of LPDDR5X memory running at 8533 MT/s. We can expect to hear more potentially at Intel's upcoming Innovation event in September, while general availability of Panther Lake is expected in late 2025 or early 2026.

In the preparation season for NVIDIA's upcoming GeForce RTX 50 Series of GPUs, codenamed "Blackwell," one power supply manufacturer accidentally leaked the power configurations of all SKUs. Seasonic operates its power supply wattage calculator, allowing users to configure their systems online and get power supply recommendations. This means that the system often gets filled with CPU/GPU SKUs to accommodate the massive variety of components. This time we have the upcoming GeForce RTX 50 series, with RTX 5050 all the way up to the top RTX 5090 GPU. Starting with the GeForce RTX 5050, this SKU is expected to carry a 100 W TDP. Its bigger brother, the RTX 5060, bumps the TDP to 170 W, 55 W higher than the previous generation "Ada Lovelace" RTX 4060.

The GeForce RTX 5070, with a 220 W TDP, is in the middle of the stack, featuring a 20 W increase over the Ada generation. For higher-end SKUs, NVIDIA prepared the GeForce RTX 5080 and RTX 5090, with 350 W and 500 W TDP, respectively. This also represents a jump in TDP from Ada generation with an increase of 30 W for RTX 5080 and 50 W for RTX 5090. Interestingly, this time NVIDIA wants to unify the power connection system of the entire family with a 16-pin 12V-2x6 connector but with an updated PCIe 6.0 CEM specification. The increase in power requirements for the "Blackwell" generation across the SKUs is interesting, and we are eager to see if the performance gains are enough to balance efficiency.

The city of Las Vegas late last year added another attraction to its town: the Sphere. The Sphere is a 1.2 million pixel outdoor display venue famous for its massive size and inner 18,600-seat auditorium. The auditorium space is a feat of its own with features like a 16x16 resolution wraparound interior LED screen, speakers with beamforming and wave field synthesis technologies, and 4D physical effects. However, we have recently found out that NVIDIA GPUs power the Sphere. And not only a handful of them, as 150 NVIDIA RTX A6000 power the Sphere and its 1.2 million outside pixels spread on 54,000 m², as well as 16 of 16K inner displays with a total output of 2.1 billion pixels. Interestingly, the 150 NVIDIA RTX A6000 have a combined output cable number of 600 DisplayPort 1.4a ports.

With each card having 48 GB of memory, that equals to 7.2 TB of GDDR6 ECC memory in the total system. With the Sphere being a $2.3 billion project, it is expected to have an infotainment system capable of driving the massive venue. And it certainly delivers on that. Only a handful of cards powers most massive media projects, but this scale is something we see for the first time in non-AI processing systems. The only scale we are used to today is massive thousand-GPU clusters used for AI processing, so seeing a different and interesting application is refreshing.

In a community blog post, Arm has announced its new Accuracy Super Resolution (ASR) upscaling technology. This open-source solution aims to transform mobile gaming by offering best-in-class upscaling capabilities for smartphones and tablets. Arm ASR addresses a critical challenge in mobile gaming: delivering high-quality graphics while managing power consumption and heat generation. By rendering games at lower resolutions and then intelligently upscaling them, Arm ASR promises to significantly boost performance without sacrificing visual quality. The technology builds upon AMD's FidelityFX Super Resolution 2 (FSR 2) and adapts it specifically for mobile devices. Arm ASR utilizes temporal upscaling, which combines information from multiple frames to produce higher-quality images from lower-resolution inputs. Even though temporal upscaling is more complicated to implement than spatial frame-by-frame upscaling, it delivers better results and gives developers more freedom.

This approach allows for more ambitious graphics while maintaining smooth gameplay. In benchmark tests using a complex scene, Arm demonstrated impressive results. Devices featuring the Arm Immortalis-G720 GPU showed substantial framerate improvements when using Arm ASR compared to native resolution rendering and Qualcomm's Game Super Resolution (GSR). Moreover, the technology helped maintain stable temperatures, preventing thermal throttling that can compromise user experience. Collaboration with MediaTek revealed significant power savings when using Arm ASR on a Dimensity 9300 handset. This translates to extended battery life for mobile gamers, addressing key concerns. Arm is releasing ASR under an MIT open-source license, encouraging widespread adoption and experimentation among developers. Below you can see the comparison of various upscalers.

According to a recent leak from Golden Pig on Weibo, AMD is gearing up to introduce the Ryzen 7 8745HS, a modified version of the existing Ryzen 7 8845HS APU. The key difference in this new chip lies in its neural processing capabilities. While the 8845HS boasts AMD's XDNA-based NPU (Neural Processing Unit), the upcoming 8745HS is rumored to have this feature disabled. Specifications for the 8745HS are expected to closely mirror its predecessor, featuring eight Zen 4 cores, 16 threads, and a configurable TDP range of 35-54 W. The chip will likely retain the Radeon 780M integrated GPU with 12 Compute Units. However, it is possible that AMD might introduce slight clock speed reductions to differentiate the new model further.

It is also worth pointing out that Hawk Point generation is not Copilot+ certified due to first-generation XDNA NPU being only 16 TOPS out of 40 TOPS required, so having an NPU doesn't help AMD advertise these processors as Copilot+ ready. The success of this new variant will largely depend on its pricing and adoption by laptop/mobile OEMs. Without the NPU, the 8745HS could offer a more budget-friendly option for users who don't require extensive local AI processing capabilities. After all, AI workloads remain a niche segment in consumer computing, and many users may find the 8745HS an attractive alternative if pricing is reduced, especially given the availability of cloud-based AI tools.

AMD today announced the signing of a definitive agreement to acquire Silo AI, the largest private AI lab in Europe, in an all-cash transaction valued at approximately $665 million. The agreement represents another significant step in the company's strategy to deliver end-to-end AI solutions based on open standards and in strong partnership with the global AI ecosystem. The Silo AI team consists of world-class AI scientists and engineers with extensive experience developing tailored AI models, platforms and solutions for leading enterprises spanning cloud, embedded and endpoint computing markets.

Silo AI CEO and co-founder Peter Sarlin will continue to lead the Silo AI team as part of the AMD Artificial Intelligence Group, reporting to AMD senior vice president Vamsi Boppana. The acquisition is expected to close in the second half of 2024.

The GeForce RTX 4070 Ti SUPER featuring the AD102 GPU, initially speculated to be an Asia-exclusive release, is now confirmed for worldwide distribution. This development comes as the GIGABYTE RTX 4070 Ti SUPER Windforce MAX OC 16 GB, the first model equipped with this configuration, has appeared on German online retailer listing, specifically at Mindfactory. Previously using AD103-275 GPU SKU for the regular GIGABYTE RTX 4070 Ti SUPER Windforce model, the "MAX" variant now comes with AD102-225 GPU SKU. Priced at over 950 Euros, it commands a premium over the standard Windforce model. The "MAX" designation not only indicates the different GPU but also boasts an official 30 MHz higher clock speed. The Max variant also employs the more robust cooling system typically found in the GeForce RTX 4080/4090 class. This is a departure from the smaller RTX 4070 Ti design used in the SUPER variant with the AD103 GPU.

While the performance and power consumption differences between the AD102 and AD103 versions may be negligible for most users, the adoption of larger cooling systems in many custom designs could prove beneficial. This enhanced thermal management might appeal to enthusiasts and overclockers looking to push their graphics cards to the limit. While the regular AD103 card measures 261x126x50 mm, the MAX model with its AD102 and beefier coolers measures 331x136x55.5 mm. Overclockers could push the performance of this card further, and we wait to see what will come out of the AD102-225 silicon on the RTX 4070 Ti SUPER platform.

Basemark announced today the release of a groundbreaking cross-platform ray tracing benchmark, GPUScore: Breaking Limit. This new benchmark is designed to evaluate the performance of the full range of ray tracing capable devices, including smartphones, tablets, laptops and high-end desktops with discrete GPUs. With support for multiple operating systems and graphics APIs, Breaking Limit provides a comprehensive performance evaluation across various platforms and devices.

As ray tracing technology becomes increasingly prevalent in consumer electronics, from high-end desktops to portable devices like laptops and smartphones, there is a critical need for a benchmark that can accurately assess and compare performance across different devices and platforms. Breaking Limit addresses this gap, providing valuable insights into how various devices handle hardware-accelerated graphics rendering. The benchmark is an essential tool for developers, manufacturers, and consumers to measure and compare the performance of real-time ray tracing rendering across different hardware and software environments reliably.

In 2016, TSMC developed and named its InFO FOWLP technology, and applied it to the A10 processor used in the iPhone 7. TrendForce points out that since then, OSAT providers have been striving to develop FOWLP and FOPLP technologies to offer more cost-effective packaging solutions.

Starting in the second quarter, chip companies like AMD have actively engaged with TSMC and OSAT providers to explore the use of FOPLP technology for chip packaging and helping drive industry interest in FOPLP. TrendForce observes that there are three main models for introducing FOPLP packaging technology: Firstly, OSAT providers transitioning from traditional methods of consumer IC packaging to FOPLP. Secondly, foundries and OSAT providers packaging AI GPUs that are transitioning 2.5D packaging from wafer level to panel level. Thirdly, panel makers who are packaging consumer ICs.

Intel has released its latest version of Arc GPU Graphics drivers for A-series discrete GPUs and Arc Graphics iGPUs, version 101.5594 Beta. The new drivers bring a fix for Citrix Workspace application on both Intel Arc A-Series and Intel Core Ultra CPU with Intel Arc GPUs that have exhibited sluggish behavior when interacting with a virtual desktop.

Unfortunately, there are no further fixes or optimizations as the previous Arc GPU Graphics drivers version 101.5593 Beta already brought optimizations for Final Fantasy XIV: Dawntrail and The First Descendant games. There are still plenty of known issues on Intel Arc A-Series graphics products including ones with games like No Man's Sky, Enshrouded, Doom Eternal and application like Topaz Video AI. There are also known issues with Intel Arc Control Studio as well with corruption of the recording file when the mouse cursor is enabled during HDR capture and issues with schedule updates for drivers.

DOWNLOAD: Intel Arc GPU Graphics Drivers 101.5594 Beta

South Korean startup Panmnesia has unveiled an interesting solution to address the memory limitations of modern GPUs. The company has developed a low-latency Compute Express Link (CXL) IP that could help expand GPU memory with external add-in card. Current GPU-accelerated applications in AI and HPC are constrained by the set amount of memory built into GPUs. With data sizes growing by 3x yearly, GPU networks must keep getting larger just to fit the application in the local memory, benefiting latency and token generation. Panmnesia's proposed approach to fix this leverages the CXL protocol to expand GPU memory capacity using PCIe-connected DRAM or even SSDs. The company has overcome significant technical hurdles, including the absence of CXL logic fabric in GPUs and the limitations of existing unified virtual memory (UVM) systems.

At the heart of Panmnesia's solution is a CXL 3.1-compliant root complex with multiple root ports and a host bridge featuring a host-managed device memory (HDM) decoder. This sophisticated system effectively tricks the GPU's memory subsystem into treating PCIe-connected memory as native system memory. Extensive testing has demonstrated impressive results. Panmnesia's CXL solution, CXL-Opt, achieved two-digit nanosecond round-trip latency, significantly outperforming both UVM and earlier CXL prototypes. In GPU kernel execution tests, CXL-Opt showed execution times up to 3.22 times faster than UVM. Older CXL memory extenders recorded around 250 nanoseconds round trip latency, with CXL-Opt potentially achieving less than 80 nanoseconds. As with CXL, the problem is usually that the memory pools add up latency and performance degrades, while these CXL extenders tend to add to the cost model as well. However, the Panmnesia CXL-Opt could find a use case, and we are waiting to see if anyone adopts this in their infrastructure.Below are some benchmarks by Panmnesia, as well as the architecture of the CXL-Opt.

TrendForce reports that a recovery in demand for general servers—coupled with an increased production share of HBM by DRAM suppliers—has led suppliers to maintain their stance on hiking prices. As a result, the ASP of DRAM in the third quarter is expected to continue rising, with an anticipated increase of 8-13%. The price of conventional DRAM is expected to rise by 5-10%, showing a slight contraction compared to the increase in the second quarter.

TrendForce notes that buyers were more conservative about restocking in the second, and inventory levels on both the supplier and buyer sides did not show significant changes. Looking ahead to the third quarter, there is still room for inventory replenishment for smartphones and CSPs, and the peak season for production is soon to commence. Consequently, it is expected that smartphones and servers will drive an increase in memory shipments in the third quarter.

A new startup emerged out of stealth mode today to power the next generation of generative AI. Etched is a company that makes an application-specific integrated circuit (ASIC) to process "Transformers." The transformer is an architecture for designing deep learning models developed by Google and is now the powerhouse behind models like OpenAI's GPT-4o in ChatGPT, Anthropic Claude, Google Gemini, and Meta's Llama family. Etched wanted to create an ASIC for processing only the transformer models, making a chip called Sohu. The claim is Sohu outperforms NVIDIA's latest and greatest by an entire order of magnitude. Where a server configuration with eight NVIDIA H100 GPU clusters pushes Llama-3 70B models at 25,000 tokens per second, and the latest eight B200 "Blackwell" GPU cluster pushes 43,000 tokens/s, the eight Sohu clusters manage to output 500,000 tokens per second.

Why is this important? Not only does the ASIC outperform Hopper by 20x and Blackwell by 10x, but it also serves so many tokens per second that it enables an entirely new fleet of AI applications requiring real-time output. The Sohu architecture is so efficient that 90% of the FLOPS can be used, while traditional GPUs boast a 30-40% FLOP utilization rate. This translates into inefficiency and waste of power, which Etched hopes to solve by building an accelerator dedicated to power transformers (the "T" in GPT) at massive scales. Given that the frontier model development costs more than one billion US dollars, and hardware costs are measured in tens of billions of US Dollars, having an accelerator dedicated to powering a specific application can help advance AI faster. AI researchers often say that "scale is all you need" (resembling the legendary "attention is all you need" paper), and Etched wants to build on that.

As Intel gears for the launch of Lunar Lake and Arrow Lake processors, the company appears to be preparing a new line of high-performance processors for gaming laptops. Recent developments suggest that the company is adapting its desktop-grade Arrow Lake-S chips for use in ultra-high-performance notebooks. The buzz began when X user @InstLatX64 spotted Intel testing a peculiar motherboard labeled "Arrow Lake Client Platform/ARL-S BGA SODIMM 2DPC." This discovery hints at the possibility of Intel packing up to 24 cores into laptop processors, eight more cores compared to the 16 cores expected in standard Arrow Lake-H mobile chips. By utilizing the full potential of Arrow Lake-S silicon in a mobile form factor, Intel aims to deliver desktop-class performance to high-end notebooks in a BGA laptop CPU.

The leaked chip would likely feature eight high-performance Lion Cove P-cores and 16 energy-efficient Skymont E-cores, along with an integrated Xe2 GPU. This configuration could provide the raw power needed for demanding games and professional applications in a portable package. However, implementing such powerful hardware in laptops presents challenges. The processors are expected to have a TDP of 45 W or 55 W, with actual power consumption potentially exceeding these figures to maintain high clock speeds. Success will depend not only on Intel's chip design but also on the cooling solutions and power delivery systems developed by laptop manufacturers. As of now, specific details about clock speeds and performance metrics remain under wraps. The test chip that surfaced showed a base frequency of 3.0 GHz, notably without AVX-512 support.

Legendary overclocker Vince Lucido, aka KINGPIN, has reportedly partnered with PNY to develop next-generation GPUs for extreme overclocking. KINGPIN, known for his unparalleled expertise in pushing hardware to its limits, revealed the partnership during a recent interview with Gamers Nexus at Computex 2024. The move comes as welcome news to enthusiasts who have been eagerly awaiting KINGPIN's next venture since EVGA's departure left a noticeable gap in the high-end GPU segment. Previously, he was the leading engineer of EVGA's high-end KINGPIN designs aimed at pushing the GPU to its limits. However, since EVGA decided to leave the GPU business, KINGPIN was looking for a new company to work on the next-generation GPU designs.

This time, the company of choice for KINGPIN is now PNY. While he has been in contact with many companies like GALAX and ASUS, he claims that it would be very crowded to work there as there are "too many cooks in the kitchen" with these companies already having in-house overclockers. He has also been talking with MSI, but the company wasn't interested in making GPUs for extreme overclocking. However, PNY has been very interested in shaking up the high-end GPU market. KINGPIN claims that there is a massive hole in the high-end GPU market, and he hopes to fill it with a collaboration with PNY. Next-generation GPU designs assisted by KINGPIN will reportedly arrive for the upcoming NVIDIA GeForce RTX 50 series of GPUs when we hope to see the legacy EVGA left to continue at PNY.

GIGABYTE TECHNOLOGY Co. Ltd, a leading manufacturer of premium gaming hardware, today launched the cutting-edge AMD Radeon PRO W7000 series workstation graphics cards, including the flagship GIGABYTE Radeon PRO W7900 Dual Slot AI TOP 48G as well as the GIGABYTE Radeon PRO W7800 AI TOP 32G. Powered by AMD RDNA 3 architecture, these graphics cards offer a massive 48 GB and 32 GB of GDDR6 memory, respectively, delivering cutting-edge performance and exceptional experiences for workstation professionals, creators and AI developers.⁠⁠

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.

According to the report from Phoronix, the upcoming Linux 6.11 kernel will introduce initial display support for the highly anticipated Intel Battlemage graphics processors. Battlemage, built on Intel's Xe2 architecture, represents the company's latest effort to challenge established players in the graphics arena. This new line of GPUs is designed to succeed the current DG2/Alchemist hardware, promising enhanced performance and improved power efficiency. The Linux 6.11 kernel will provide the fundamental capability to drive displays connected to Battlemage GPUs. While this initial support is a crucial first step, it lays the groundwork for more comprehensive functionality in future updates. Linux users and developers can look forward to testing and providing feedback on these new graphics solutions.

Intel's focus on power efficiency is evident in the features accompanying Battlemage support. The kernel will introduce Panel Replay functionality, a technology aimed at reducing display power consumption. This aligns with the growing demand for energy-efficient computing solutions, particularly in mobile and laptop segments. However, the work is far from complete. Intel's Linux graphics driver team continues to refine and optimize Battlemage support, with the goal of delivering a seamless experience by the time these GPUs hit the market later this year. The introduction of Battlemage support not only expands options for Linux users but also intensifies competition in the GPU market, potentially driving innovation across the industry. With promises of up to 1.5x over the previous generation Arc GPUs, we are in for a decent upgrade this year.

PC enthusiasts who crave powerful graphics in compact systems have a new option from Zephyr. The Chinese manufacturer has launched the world's first ITX-sized GeForce RTX 4070 graphics card. Dubbed the Sakura Blizzard, this GPU packs NVIDIA's AD104-250 chip and 12 GB of GDDR6X memory into a footprint of just 172 x 123 x 42 mm. While slightly taller than standard PCIe brackets, the two-slot cooler should fit most Mini-ITX cases. The card's cute pink shroud and solitary cooling fan give it a unique aesthetic. But don't let the pink looks fool you - Zephyr claims this compact powerhouse can keep the GPU and memory up to 10°C cooler than some dual-fan RTX 4070 designs, which needs to be taken with a grain of salt. Thermal testing videos show the fan spinning at 2,400 RPM to maintain GPU temperatures around 73°C under load in a 25°C room. Meanwhile, synthetic benchmarks reportedly demonstrate no performance compromises versus full-sized RTX 4070 implementations.

Zephyr's initial production run has already sold out in China. However, a second batch is slated for mid-July availability to meet the apparent higher demand for small form factor RTX 40-series GPUs. The launch comes just weeks after NVIDIA unveiled new "SFF-ready" design guidelines at Computex 2024. As the power-hungry RTX 40 lineup hit the market, many voiced concerns over the cards' ever-growing dimensions. NVIDIA's renewed SFF PC focus signal options like the Sakura Blizzard could become more common. For space-constrained enthusiasts, having top-tier GPU muscle in a properly-cooled Mini-ITX card is a big win. Zephyr's ITX-sized RTX 4070 shows powerful things can come in small packages, and we hope more manufacturers follow this philosophy.

Stability AI, a maker of various generative AI models and the company behind text-to-image Stable Diffusion models, has released its latest Stable Diffusion 3 (SD3) Medium AI model. Running on two billion dense parameters, the SD3 Medium is the company's most advanced text-to-image model to date. It boasts features like generating highly realistic and detailed images across a wide range of styles and compositions. It demonstrates capabilities in handling intricate prompts that involve spatial reasoning, actions, and diverse artistic directions. The model's innovative architecture, including the 16-channel variational autoencoder (VAE), allows it to overcome common challenges faced by other models, such as accurately rendering realistic human faces and hands.

Additionally, it achieves exceptional text quality, with precise letter formation, kerning, and spacing, thanks to the Diffusion Transformer architecture. Notably, the model is resource-efficient, capable of running smoothly on consumer-grade GPUs without compromising performance due to its low VRAM footprint. Furthermore, it exhibits impressive fine-tuning abilities, allowing it to absorb and replicate nuanced details from small datasets, making it highly customizable for specific use cases that users may have. Being an open-weight model, it is available for download on HuggingFace, and it has libraries optimized for both NVIDIA's TensorRT (all modern NVIDIA GPUs) and AMD Radeon/Instinct GPUs.

The race is on for memory manufacturers to bring the next generation GDDR7 graphics memory into mass production. While rivals Samsung and Micron are aiming to have GDDR7 chips available in Q4 of 2024, South Korean semiconductor giant SK Hynix revealed at Computex 2024 that it won't kick off mass production until the first quarter of 2025. GDDR7 is the upcoming JEDEC standard for high-performance graphics memory, succeeding the current GDDR6 and GDDR6X specifications. The new tech promises significantly increased bandwidth and capacities to feed the appetites of next-wave GPUs and AI accelerators. At its Computex booth, SK Hynix showed off engineering samples of its forthcoming GDDR7 chips, with plans for both 16 Gb and 24 Gb densities.

The company is targeting blazing-fast 40 Gbps data transfer rates with its GDDR7 offerings, outpacing the 32 Gbps rates its competitors are starting with on 16 Gb parts. If realized, higher speeds could give SK Hynix an edge, at least initially. While trailing a quarter or two behind Micron and Samsung isn't ideal, SK Hynix claims having working samples now validates its design and allows partners to begin testing and qualification. Mass production timing for standardized memories also doesn't necessarily indicate a company is "late" - it simply means another vendor secured an earlier production window with a specific customer. The GDDR7 transition is critical for SK Hynix and others, given the insatiable demand for high-bandwidth memory to power AI, graphics, and other data-intensive workloads. Hitting its stated Q1 2025 mass production target could ensure SK Hynix doesn't fall too far behind in the high-stakes GDDR7 race, with faster and higher-density chips to potentially follow shortly after volume ramp.

At Computex 2024, Colorful has followed the success of their popular COLORFIRE MEOW Series gaming desktops. The company has unleashed two fierce new gaming laptops - the COLORFIRE MEOW R15 and R16. These portable laptops are adorned with designs inspired by Bobi, COLORFIRE's beloved office cat. From paw print patterns to a sleek orange tabby color scheme, the laptops are littered with feline flair. Under the cat-inspired hood, the MEOW laptops pack a serious punch. They are equipped with the latest AMD Ryzen 8000 Series processors tailored for gaming, content creation, and entertainment. AMD Ryzen AI technology provides a boost to artificial intelligence capabilities. Both run on the AMD Ryzen 7 8845HS processor with eight cores, 16 threads, and a 5.1 GHz maximum boost clock. The Ryzen AI engine delivers up to 39 TOPS for accelerated AI performance. For graphics, the laptops feature up to an NVIDIA GeForce RTX 4070 GPU with 140 W of power.

The R15 has a 15.6-inch 2560x1440 IPS panel with 100% sRGB color and 165 Hz refresh rate. The R16 goes even bigger with a 16-inch 2560x1600 IPS display, 100% sRGB, and 240 Hz refresh. AMD FreeSync is there as well to ensure a tear-free experience. Colorful has outfitted the laptops with dual turbo fans, a 6-heatpipe solution, and performance control software to tame the heat from the powerful components. When it's time to pounce on the latest games, these MEOW laptops have the fierce cooling to keep their claws out.

At Computex 2024, Colorful has introduced a new PC series of premium componentry for games, called iGame Ultra Series. Carrying a premium white aesthetics, the Ultra Series serves as clean and elegant component for PC builds. At the heart of the Ultra Series is the GeForce RTX 4080 SUPER graphics card. The card has a triple-fan cooling solution with vapor chamber technology that keeps temperatures under control. Powering this GPU is the iGame Z790D5 Ultra motherboard. Supporting the latest Intel 14th Gen CPUs and up to 192 GB of DDR5-7800 RAM. An array of ultra-fast PCIe 5.0 and 4.0 M.2 slots allow multiple high-speed SSDs for ridiculous storage performance. Keeping all of these elite components running is the 850 W iGame P850G Ultra W 80 Plus Gold power supply. Its modular design with dedicated PCIe 5.0 connectors ensures clean cable management and stable power delivery. For cooling the CPU, there is the iGame LQ360 Ultra W all-in-one liquid cooler, which sports a massive 360 mm radiator and a trio of high-static pressure fans. All of these premium components are housed in the iGame C23A Ultra W full-tower chassis. With a spacious dual-compartment layout, front USB-C port, and preinstalled vertical GPU bracket, it offers both incredible expandability and sleek aesthetics.