The acceleration within the AI space has sparked a revolution in the PC segment. We have seen major chipmakers embark on the "AI PC" journey, which involves dedicated client designs that mix Windows applications with dedicated and local AI subsystems, and then we have dedicated boxes for AI, which are made primarily for local acceleration of models and agents. Last year, NVIDIA introduced its DGX Spark PC, the first PC of its kind to tackle LLMs and Agents in what is dubbed to be an "AI Supercomputer". Soon after the launch, other manufacturers started to label their own Mini PCs […]

Read full article at https://wccftech.com/review/nvidia-dgx-spark-ai-supercomputer-impressions-the-future-of-ai-computing-in-tiny-boxes/

According to the August Steam Hardware Survey, AMD’s Radeon RX 9070 XT has now become the company’s most widely‑used discrete GPU on Steam. The card holds a 1.46% share among surveyed systems, placing it 24th overall — just ahead of Nvidia’s GeForce RTX 3080 and directly behind the GeForce GTX 1660 Super.

Valve’s latest figures (via VideoCardz) show steady month‑to‑month growth for the RX 9070 XT, rising from 1.33% in May to 1.35% in June, 1.40% in July and 1.46% in August. AMD’s next RDNA 4 entry on the list is the Radeon RX 9060 XT at 0.80%.

Nvidia continues to dominate the upper portion of the chart with its current GeForce RTX 50‑series lineup. The GeForce RTX 5070 sits in third place overall at 3.77%, followed by the RTX 5060 at 3.16% and the RTX 5060 Ti at 2.55%. Even the higher‑end RTX 5080 maintains a 1.71% share.

Intel’s results remain unchanged, with integrated GPUs making up the bulk of its presence. Intel Iris Xe graphics is the most common at 1.74%, while dedicated Arc GPUs still do not appear in the survey.

KitGuru Says: The Radeon RX 9070 XT continues to climb, but Nvidia’s RTX 50‑series cards still hold a clear lead across the Steam user base.

The post AMD Radeon RX 9070 XT becomes company’s top discrete GPU on Steam first appeared on KitGuru.

Desktop graphics cards have been designed to be compatible with mid-tower or full-tower chassis, and to make them fit and operate in a rackmount server requires immense creativity, not to mention the risk that comes with it. Since these server types feature non-standard ATX PCIe power connectors, a regular GPU wouldn’t work, but here’s the beauty of modding: you can make non-compatible hardware work. Using their soldering skills, a group of individuals added some Molex-style cables to the graphics card, making it work with the rackmount server. Instead of taking the boring way and purchasing non-standard cables, the modders decided to […]

Read full article at https://wccftech.com/nvidia-rtx-gpu-soldered-molex-server-mod/

The new generation of data center systems is increasing efficiency with smarter traffic control instead of just more processor cycles.

We learned more about the new 160GB to 480GB LPDDR5X Intel Crescent Island GPU focused on memory capacity at Hot Chips 2026

The post Intel Crescent Island 160GB to 480GB LPDDR5X AI GPU at Hot Chips 2026 appeared first on ServeTheHome.

mVolt+ is a relatively new overclocking utility that has been receiving rapid updates since launch, with the latest version adding advanced power‑channel and voltage controls for Nvidia GeForce RTX 50 series GPUs – options not currently available in tools like MSI Afterburner.

As spotted by Videocardz, mVolt+ Release v0.36 introduces separate power‑channel limits for the core and memory, alongside voltage controls across the Core, XBAR, SYS and Video domains. These additions allow certain GPUs to exceed total power limits defined in their VBIOS, enabling behaviour previously restricted to XOC BIOS or shunt‑mod configurations.

Over on Reddit, users have already begun sharing results. One RTX 5090 owner posted a 3DMark Steel Nomad run showing the card pulling close to 700W under load. Another reported core clocks between 3.15GHz and 3.20GHz, with an XBAR frequency around 2.8GHz.

Despite the expanded controls, mVolt+ cannot bypass physical hardware limits. The developer warns that extreme voltage tuning may cause instability and void warranties.

KitGuru says: While mVolt+ unlocks impressive benchmark scores, the massive increase in power draw makes it a bit too risky to run permanently.

The post mVolt+ adds advanced voltage and power controls for RTX 50 GPUs first appeared on KitGuru.

AMD has published the firmware required for its upcoming GFX 11.7-based RDNA 4m graphics. AMD's Upcoming RDNA 4m Graphics Take Another Step Toward Launch as Firmware Support Goes Live AMD is making some significant progress in preparing its next-gen RDNA 4m-based products. It has now added the firmware binaries for the RDNA 4m graphics architecture, providing a major clue that its next-generation APUs are getting closer to release. The firmware has been upstreamed to the "linux-firmware" repository, where Linux distributions can package it alongside the required AMDGPU drivers. The RDNA 4m platform has attracted our attention because it doesn't follow […]

Read full article at https://wccftech.com/amd-publishes-rdna-4m-firmware-ahead-of-launch/

Gaming laptops give users the flexibility to switch between “Eco” mode to exclusively run the machine on the integrated GPU to get better battery life, with the option to switch to “Hybrid” or “Dedicated” modes for gaming purposes. However, one user noticed that in his quest to run on Eco mode entirely on his new ASUS ROG Zephyrus G14, Windows ended up uninstalling the dGPU driver on its own without informing the user, leaving “Microsoft Basic Display Adapter” in its place. A gaming laptop owner discovered that Windows can uninstall dGPU drivers in 15 to 30 days when running strictly […]

Read full article at https://wccftech.com/windows-deletes-dgpu-driver-gaming-laptop-eco-mode/

The AI buildout shows no signs of slowing. And with hundreds of billions of dollars a year going into data centers and GPUs, compute has become the single biggest cost for anyone building AI products. But for all that spending, there still isn’t a straightforward way to put a price on compute — or for firms to hedge their exposure when the price changes.  Silicon Data […]

After NVIDIA and CEO Jensen Huang announced their latest initiative last week, which now turns the firm's GPU into an asset class similar to financial securities, Trump advisor David Sacks discussed the biggest risk to the plan and the ongoing AI buildout. NVIDIA's latest announcement will also see the firm backstop the GPUs by providing residual support, Huang outlined. In his appearance on the All In Podcast, Sacks outlined that an oversupply of compute remained the biggest risk to the AI buildout. Sacks Says Opposition To AI Data Centers Might Work In Favor Of The Ongoing Buildout With big tech's […]

Read full article at https://wccftech.com/us-tech-advisor-david-sacks-calls-the-biggest-risk-to-nvidias-500-billion-ai-financing-plan-a-dark-gpu-glut/

MSI has launched its new UK Store, bringing gaming laptops, handhelds, components and peripherals together under an official storefront, along with exclusive bundles and limited‑edition hardware that won't be available elsewhere.

The site offers an easier way to browse MSI’s full product range, from graphics cards and motherboards through to monitors and accessories. Early sign‑ups receive a £20 voucher for orders over £150, along with free delivery for new members.

To mark the launch, MSI is running a “Cheat Code” campaign between the 11th and 18th of August. Each weekday, a £1 upgrade code will be hidden somewhere on the store, with the first shopper to find the code and complete checkout securing the corresponding item at an incredibly cheap price. Hardware drops include graphics cards, liquid‑cooling bundles, QD‑OLED monitors, PC cases with power supplies and networking gear. The hunt resets daily, giving visitors multiple chances to take part.

Discuss on our Facebook page, HERE.

KitGuru Says: MSI’s launch campaign adds a bit of gamification to the new storefront. Will you be hunting for one of the codes?

The post MSI launches official UK store with £1 hardware first appeared on KitGuru.

GMKtec has been making Mini PCs for years now. They are based out of the technology hub of Shenzhen, China, and have maintained strategic partnerships with AMD, Intel & Microsoft, and currently offer a wide array of solutions, all the way from entry-level to enthusiast-grade designs based on mobility chips from AMD & Intel. The latest offering from GMKtec that targets the high-end segment with Intel's Core Ultra Series 3 SoCs. For this, GMKtec is introducing its EVO T2S, which is a follow-up to the EVO T1 we tested last year with the Core Ultra 9 285H chip. The new […]

Read full article at https://wccftech.com/review/gmktec-nucbox-evo-t2s-intel-core-ultra-x7-358h-mini-pc-review/

The appropriate depreciation rate for NVIDIA GPUs has consistently ranked as one of the biggest unknowns of the ongoing AI build out, with proponents continuing to argue in favor of a longer shelf life, while detractors contend that a 3-year useful life estimate is most appropriate. The latest hourly rental rate figures for NVIDIA GPUs, however, appear to bolster the point of view espoused by the proponents of longer useful life estimates, which bodes well for the income statements of AI hyperscalers. Older NVIDIA GPUs, such as the A100s and H100s, command a higher rental rate today than what they […]

Read full article at https://wccftech.com/what-gpu-depreciation-nvidia-a100-h100-and-h200-gpu-rental-rates-continue-to-price-upwards-as-b200s-hover-at-just-shy-of-6-hour/

The team Red has started laying the groundwork for the RDNA 5 lineup on Linux by introducing initial support for DCN6. Early AMDGPU Patches Introduce Display Core Next 6, Tied to the Upcoming GFX13 Graphics Architecture AMD has taken one more step toward enabling support for its next-generation RDNA architecture on Linux. A new batch of AMDGPU patches has been submitted to the Linux Kernel mailing list and has begun laying the foundation for Display Core Next 6, aka DCN6, which is the display engine expected to power the RDNA 5 graphics chips. These patches are part of AMD's regular […]

Read full article at https://wccftech.com/amd-prepares-linux-driver-support-for-rdna-5-gpus-with-new-dcn6-display-engine-patches/

The higher video memory options for consumer GPUs are only available with products like NVIDIA’s RTX 5090, but one service provider is offering VRAM upgrades for those who don’t wish to make another upgrade and don’t know where to begin to get more life out of their graphics card. By increasing the total memory, gamers can not only play the newest AAA titles without running into low-memory issues but also run LLMs while they’re at it. GPU memory upgrade services have already demonstrated an RTX 2080 Ti with 11GB of VRAM now shipping with 22GB GPU Solutions offers unique services […]

Read full article at https://wccftech.com/modders-offer-gpu-memory-upgrades-for-gaming-and-ai/

ASRock has launched the AMD Radeon RX 9050 Challenger 8GB graphics card. Based on AMD's latest RDNA 4 architecture, the new dual-fan graphics card arrives with 8GB of GDDR6 memory and is heading to system integrators and retail customers across LATAM and Asia Pacific.

Built on the RDNA 4 architecture, the AMD Radeon RX 9050 features up to 16 CUs, 16 ray accelerators based on 3rd Gen Raytracing hardware, and 32 AI accelerators. Supported software features include AMD HYPR-RX and AMD FSR technology, alongside modern hardware standards like DisplayPort 2.1a and PCI Express 5.0 compatibility. ASRock sets the Boost Clock on its Challenger model to up to 2600MHz, while the game clock is set to 1920MHz. That translates into 10.6TFLOPS of FP32 computing performance, which is just slightly above what you can expect from a PS5.

The 8GB of GDDR6 memory is clocked at 18Gbps over a 128-bit memory interface. Moreover, it has 32MB of Infinity Cache. ASRock recommends a 450W PSU for a system featuring its new card, but the card's rated TBP is only 92W.

Cooling is handled by a dual-fan thermal solution incorporating ASRock's Striped Axial Fan design (with 0db operation) and Ultra-fit Heatpipes. Additional hardware features include a metal backplate for structural rigidity and integrated LED indicators.

KitGuru says: ASRock's new RX 9050 Challenger 8GB gives mainstream builders in LATAM and Asia Pacific an affordable entry point into AMD's latest RDNA 4 architecture. Do you think this card should also be available in other regions?

The post ASRock introduces Radeon RX 9050 Challenger 8GB with dual-fan cooler first appeared on KitGuru.

Primate Labs has officially released Geekbench 7, introducing updated CPU and GPU workloads alongside redesigned multi-core testing. The new cross-platform benchmark suite is available today across Android, iOS, Windows, macOS, and Linux platforms.

Geekbench 7's updated CPU workloads focus on real-world media tasks, including AV1 video encoding for screen-sharing, Opus audio compression, and Whisper speech recognition for live captions. Primate Labs also added a Game Physics workload built on the Jolt Physics engine while expanding image format support in the Photo Library to include JPEG XL and DNG. Multi-core testing has been restructured so workloads only run multi-threaded if corresponding real-world applications do, excluding single-threaded tasks like the HTML5 Browser test.

GPU benchmarks now place greater emphasis on machine learning and creation tasks, adding face tracking, super-resolution upscaling, and background blurring alongside RAW processing, path tracing, and fluid simulation. In response to user demand, CUDA joins OpenCL, Vulkan, and Metal as a natively supported GPU API. Primate Labs updated the underlying datasets across tests such as File Compression and PDF Viewer to support larger, more demanding file processing.

KitGuru says: Geekbench 7 brings welcome updates to reflect modern workloads, particularly with the addition of native CUDA support and a more realistic multi-core scoring approach. Have you already tried Geekbench 7?

The post Geekbench 7 arrives with redesigned multi-core testing and CUDA support first appeared on KitGuru.

AMD has announced new additions to its data‑centre portfolio at Advancing AI 2026, introducing both the 6th Gen EPYC 9006 Series server CPUs and the Instinct MI400 Series GPUs, aimed at next-generation AI infrastructure.

The EPYC 9006 Series is described by AMD as a workload‑optimised CPU lineup built to support agentic AI, where CPUs take on a larger share of work by running agents, coordinating tasks and managing data flows. AMD claims the new generation is designed to handle this increased load through higher core counts, expanded memory bandwidth and support for next‑generation I/O standards such as PCIe Gen 6 and MRDIMM.

The EPYC 9006 Series consists of four CPU families. The SP7 tier is the maximum‑density option, with core counts up to 256 cores and 512 threads, along with high‑frequency variants reaching up to 5GHz. SP8 models focus on efficiency, offering up to 128 core configurations designed for enterprise, edge and power‑constrained deployments. For HPC and memory‑sensitive workloads, AMD points to the SP7X variants, which it says offer triple the L3 cache per core compared to SP7. The LP tier serves as an AI host‑node CPU, using LPDDR memory and higher frequencies to keep accelerators supplied with data in multi-GPU racks.

On the GPU side, AMD introduced the Instinct MI400 Series, split between the MI455X for frontier‑scale AI deployments and the MI430X for sovereign AI and HPC. The MI455X is AMD's highest‑end accelerator for large‑model training, fine‑tuning and high‑volume inference, while the MI430X is aimed at workloads requiring numerical accuracy. AMD says the MI430X can deliver up to 288 TFLOPS of FP64 performance, alongside HBM4 memory and high‑bandwidth interconnects to support larger models.

The MI400 Series also incorporates security and system‑level features for sensitive AI and HPC environments, including SecureBoot, encrypted GPU‑to‑GPU links and hardware‑based protections. These GPUs support AMD’s ROCm software stack for AI development, which is also receiving updates this week through ROCm.ai – a new AI‑native developer environment that adds a unified CLI, AMD‑authored guidance for coding assistants and AI‑driven workload optimisation.

Finally, AMD is also expanding into physical AI and robotics, introducing new Ryzen AI Embedded X100 processors and Kria AI system‑on‑modules for autonomous machines and industrial systems. These platforms combine Zen 5 CPU cores, integrated RDNA graphics and an NPU, with shared access to memory to reduce latency in real‑time workloads.

KitGuru Says: What do you think of AMD's push into the AI space? 

The post AMD expands AI portfolio with EPYC 9006, Instinct MI400 and new physical‑AI hardware first appeared on KitGuru.

The new variant is said to have been cancelled as the RTX 3060 12 GB has already entered the market. GeForce RTX 5050 9 GB GPU May Never Launch As NVIDIA Has Reportedly Cancelled or Delayed it The budget segment doesn't have many options from the current-gen NVIDIA and AMD series, which is why NVIDIA wanted to expand the RTX 50 series lineup with another variant of the GeForce RTX 5050. If you have been following reports surrounding the RTX 5050 with 9 GB VRAM, you might be aware that the GPU was supposedly delayed, but was still planned for […]

Read full article at https://wccftech.com/nvidia-reportedly-kills-rtx-5050-9-gb-before-launch-as-its-own-rtx-3060-12-gb-makes-it-pointless/