Topic / Trend Rising

AI Infrastructure and Hardware

The demand for AI is driving growth in the semiconductor industry, memory solutions, and data center infrastructure. Companies are investing in advanced packaging, HBM, and new chip architectures to support AI workloads.

Detected: 2026-01-31 · Updated: 2026-02-15

Related Coverage

2026-02-15 LocalLLaMA

AdaLLM: NVFP4-first inference on RTX 4090 with FP8 cache

AdaLLM is a new runtime focused on accelerated NVFP4 inference on Ada Lovelace GPUs, such as the RTX 4090. It implements an FP8 KV cache and a custom FP8 decode kernel, avoiding FP16 fallback. It supports models like Qwen3 and Gemma3, offering perfor...

#Hardware #LLM On-Premise #DevOps
2026-02-14 DigiTimes

Taiwan's January exports jump nearly 70% on AI server boom

Taiwan's January exports jumped nearly 70%, driven by strong demand for AI servers. This growth fuels optimism for 2026 and strengthens investments in technological defense. The Taiwanese tech sector benefits from the increasing demand for artificial...

#Hardware #LLM On-Premise #DevOps
2026-02-14 Tom's Hardware

President's Day GPU bonanza: RTX 5060 Ti 8GB back to $379 MSRP

For President's Day, MSI is offering the GeForce RTX 5060 Ti 8GB Ventus 2X OC Plus at the MSRP of $379. The deal also includes the possibility to score a Blackwell GPU at MSRP, and a free 650W power supply.

#Hardware #LLM On-Premise #DevOps
2026-02-13 Phoronix

NVIDIA Posts Open-Source Nouveau GSP Driver Support For GA100

NVIDIA has released an open-source update for the Nouveau driver, focusing on supporting the GA100 GPU through the GPU System Processor (GSP). This move aims to improve the compatibility and performance of NVIDIA GPUs with open-source drivers.

#Hardware #LLM On-Premise #DevOps
2026-02-13 ServeTheHome

OpenAI GPT-5.3 Achieves 1000 Tokens/Second on Cerebras Chips

OpenAI's GPT-5.3-Codex-Spark model has been optimized to run on Cerebras WSE-3 processors, achieving an inference speed of over 1000 tokens per second. This performance opens new perspectives for applications requiring fast, low-latency responses.

#LLM On-Premise #DevOps
2026-02-13 Tom's Hardware

Radeon RX 9060 XT 8GB defies the AI crunch and drops back to $299 MSRP

The Radeon RX 9060 XT 8GB, based on Navi 44 architecture, is repositioning itself in the market with an MSRP of $299. This move aims to provide an affordable solution for budget gamers, amidst increasing demand and high costs for AI-focused GPUs.

#Hardware #Fine-Tuning
2026-02-13 404 Media

Waymo Is Getting DoorDashers to Close Doors on Self Driving Cars

Waymo, Google's autonomous vehicle company, has launched a pilot program with DoorDash. Dashers are paid to close the doors of self-driving cars left open by passengers. The initiative highlights the need for human intervention even in the most autom...

#LLM On-Premise #DevOps
2026-02-13 LocalLLaMA

Deepseek testing a new model: focus on reading comprehension

Deepseek, a Chinese group active in the development of large language models (LLM), has announced that it is testing a new model. Preliminary benchmarks focus on reading comprehension skills, with results showing variable performance across different...

#LLM On-Premise #DevOps
2026-02-13 DigiTimes

Quanta boosts Thailand investment to expand AI server capacity

Quanta Computer is expanding its AI server production in Thailand, increasing investments to meet growing demand. The initiative aims to strengthen production capacity in the artificial intelligence sector, with a focus on dedicated servers.

#LLM On-Premise #Fine-Tuning #DevOps
2026-02-13 DigiTimes

AUO to hire 1,000 in 2026 as AI expands display, smart mobility push

Display manufacturer AUO plans to hire 1,000 people by 2026. The expansion is driven by increasing demand for AI solutions in the display and smart mobility sectors. The company aims to strengthen its presence in these growing markets.

#LLM On-Premise #DevOps
2026-02-13 DigiTimes

Samsung and SK Hynix: HBM and AI focus at SEMICON Korea

At SEMICON Korea, Samsung highlighted advancements in HBM memory with hybrid bonding technology. SK Hynix, on the other hand, emphasized the integration of artificial intelligence at the platform level, signaling a paradigm shift in the approach to t...

#LLM On-Premise #DevOps
2026-02-13 DigiTimes

AI chip spending nears $1tn tipping point

Global spending on AI chips is rapidly increasing, approaching a tipping point of $1 trillion. This surge reflects the growing demand for computing power for AI applications.

#Hardware #LLM On-Premise #DevOps
2026-02-13 DigiTimes

Taiwan: debate over scientific research metrics ignited

Minister Cheng-wen Wu's remarks on the importance of scientific research evaluation metrics in Taiwan have sparked a debate in the academic community. The discussion revolves around how to effectively measure the impact and quality of research, and w...

2026-02-13 TechWire Asia

Visa: AI commerce trust gap widens in Asia Pacific at checkout

A Visa survey reveals that despite widespread adoption of AI for product discovery in Asia Pacific, nearly half of consumers hesitate to use it for payments. More experienced and affluent consumers are the most skeptical, fearing data security and al...

#LLM On-Premise #DevOps
2026-02-12 Ars Technica AI

OpenAI sidesteps Nvidia with GPT-5.3-Codex-Spark coding model on Cerebras

OpenAI released GPT-5.3-Codex-Spark, its first production AI model to run on non-Nvidia hardware, deploying on Cerebras chips. The model delivers code at over 1,000 tokens per second, roughly 15 times faster than its predecessor. Access is available ...

#Hardware #LLM On-Premise #DevOps
2026-02-12 TechCrunch AI

SpaceX Eyes Moonbase Alpha for Deep Space AI Inference

Elon Musk envisions a lunar base, dubbed Moonbase Alpha, equipped with a mass driver system for launching AI satellites into deep space. The goal is to accelerate AI development and space exploration.

#LLM On-Premise #DevOps
2026-02-12 PyTorch Blog

PyTorch accelerates type checking with Pyrefly

PyTorch has adopted Pyrefly for type checking, achieving a 10x speed increase compared to MyPy. The migration simplifies configuration, ensures consistency across development environments, and improves code quality with advanced typing features. Cont...

2026-02-12 Tom's Hardware

Nvidia DGX Spark update cuts idle power by 32% or more

Nvidia DGX Spark update introduces hot-plug detection on ConnectX NIC, optimizing the energy efficiency of AI workstations and reducing idle power consumption by 32% or more. A step forward for more efficient AI workstations.

#Hardware #LLM On-Premise #DevOps
2026-02-12 TechCrunch AI

OpenAI's Codex: new version powered by a dedicated chip

OpenAI announced a new version of its Codex coding tool, highlighting it as a milestone in its relationship with a chipmaker. No details were provided on the chip's technical specifications or the performance improvements achieved.

#Hardware #Fine-Tuning
2026-02-12 Phoronix

Ubuntu 24.04.4 LTS Now Available With Linux 6.17 HWE Kernel

Canonical released Ubuntu 24.04.4 LTS, the newest point release to the Noble Numbat series. This release includes security updates, bug fixes, and the Linux 6.17 HWE (Hardware Enablement) kernel. The goal is to provide a more stable and up-to-date ba...

#Hardware #LLM On-Premise #DevOps
2026-02-12 Tom's Hardware

Chinese investment in AI and data centers accelerates

An analysis highlights how China's push for technological self-sufficiency is fueling massive investments in the artificial intelligence and data center sectors. This trend underscores the growing strategic importance of these sectors for the Chinese...

#LLM On-Premise #DevOps
2026-02-12 Tom's Hardware

U.S. Lawmakers Demand Sales Ban on Chipmaking Tools to China

A bipartisan group of U.S. lawmakers is demanding a ban on the export of chipmaking tools to China, specifically targeting ASML's lithography machines. The aim is to limit China's ability to develop advanced semiconductors.

2026-02-12 The Next Web

Swedish pet insurtech Lassie raises $75M Series C

Lassie, a Swedish insurtech startup for pets, has announced a $75 million Series C funding round after reaching $100 million in annual recurring revenue (ARR). The company focuses on prevention, providing owners with tools to take care of their pets.

2026-02-12 DigiTimes

Samsung reclaims AI crown with world-first HBM4 mass production

Samsung announces the start of mass production of HBM4 memories, a world first that could redefine performance in artificial intelligence and high-performance computing. This move consolidates Samsung's position in the high-speed memory market.

#Hardware #LLM On-Premise #DevOps
2026-02-12 AI News

State-sponsored hackers exploit AI for advanced cyberattacks

State-sponsored hackers are exploiting AI models like Gemini to refine phishing attacks and develop malware. Groups from Iran, North Korea, China, and Russia are leveraging AI for reconnaissance, social engineering, and malware development, increasin...

2026-02-12 DigiTimes

Mistral to invest EUR1.2bn in Swedish AI data centres

Mistral AI plans to invest EUR1.2 billion in Sweden to expand its European compute capacity through new data centers dedicated to artificial intelligence. The initiative aims to strengthen Mistral's presence in the European AI landscape.

2026-02-12 DigiTimes

Taiwan PCB industry unites at Apex Expo for US advanced packaging push

Taiwanese printed circuit board (PCB) manufacturers are coordinating to support the development of advanced packaging solutions in the United States. The initiative aims to strengthen the supply chain and respond to the growing demand for advanced te...

#LLM On-Premise #DevOps
2026-02-12 DigiTimes

Taiwan charts course toward defense autonomy with NT$400 billion push

Taiwan has announced an investment of NT$400 billion (approximately US$12.5 billion) to strengthen its autonomy in the defense sector. The initiative aims to promote the development of local technologies and capabilities, reducing dependence on forei...

#LLM On-Premise #DevOps
2026-02-12 DigiTimes

Musk announces xAI reorganization after merger with SpaceX

Elon Musk has announced a reorganization of xAI following its merger with SpaceX. Specific details of the reorganization have not been disclosed, but the move suggests a closer integration between the two companies.

#LLM On-Premise #DevOps
2026-02-12 DigiTimes

SEMI: AI and HBM lift 2025 silicio wafer shipments, revenue still dips

According to SEMI, silicio wafer shipments are expected to increase in 2025 due to demand for AI and HBM (High Bandwidth Memory). Despite the projected growth, overall industry revenue will remain below previous peaks. The article analyzes trends in ...

#LLM On-Premise #DevOps
2026-02-12 DigiTimes

Taiwan carriers post January gains, shift spending toward AI and cloud

Taiwanese carriers report growth in January and plan to increase investments in cloud infrastructure and artificial intelligence solutions. This transition reflects a global trend towards adopting advanced technologies to improve services and optimiz...

#LLM On-Premise #DevOps
2026-02-12 DigiTimes

China claims five places in 2025 global OSAT top 10

China is poised to solidify its position in the Outsourced Semiconductor Assembly and Test (OSAT) sector. Projections indicate that five Chinese companies will rank among the top ten globally by 2025, a clear sign of the country's growing influence i...

#LLM On-Premise #DevOps
2026-02-12 Phoronix

Linux 7.0: Graphics Drivers Updated with AMD and Intel Xe Support

Linux kernel 7.0 introduces significant updates to DRM (Direct Rendering Manager) graphics drivers, featuring enhancements for AMD hardware and SR-IOV support for Intel Xe. Also included are improvements to "accel" drivers for AI accelerators like NP...

#Hardware #LLM On-Premise #DevOps
2026-02-11 TechCrunch AI

AI inference startup Modal Labs in talks to raise at $2.5B valuation

AI inference startup Modal Labs is in talks for a new funding round led by General Catalyst, potentially valuing the company at $2.5 billion. The four-year-old company is rapidly establishing itself in the artificial intelligence landscape.

#LLM On-Premise #DevOps
2026-02-11 Tom's Hardware

Ryzen 7 9800X3D: PBO settings match pricier 9850X3D in gaming

Testing reveals that the Ryzen 7 9800X3D, with simple PBO (Precision Boost Overdrive) settings, can match the gaming performance of the pricier Ryzen 7 9850X3D. The higher clock speed of the latter doesn't provide a significant advantage in gaming sc...

#Hardware
2026-02-11 The Register AI

Only 20% of European Datacenters are AI-Ready

A report highlights that only 20% of data centers in Europe and the Middle East are equipped to handle artificial intelligence workloads. Skills shortages and grid bottlenecks threaten to stall capacity expansion.

#LLM On-Premise #DevOps
2026-02-11 Tom's Hardware

Microsoft turns to superconductors for AI data centers

Microsoft is exploring the use of superconducting cables to power its AI data centers. This technology promises to reduce power losses and heat emissions, improving overall energy efficiency.

#LLM On-Premise #DevOps
2026-02-11 DigiTimes

GPTC management shakeup raises concerns over TSMC orders

A management shakeup at GPTC raises questions about the stability of orders to key suppliers such as TSMC, ASE, and Micron. The reorganization could have repercussions on the supply chain of essential components for artificial intelligence.

#Hardware #LLM On-Premise #DevOps
2026-02-11 The Register AI

As OpenAI and Claude fight over ads, Google says ‘show me the money’

As OpenAI walks the advertising tightrope to balance revenue gains against credibility and safety, ad kingpin Google is roaring ahead to use AI to improve its advertising products. Google isn't showing ads in Gemini, but AI Mode is fair game.

#LLM On-Premise #DevOps
2026-02-11 DigiTimes

TSMC Japan expansion mirrors broader Taiwan tech retreat from China

TSMC's expansion in Japan may reflect a broader trend of Taiwanese tech companies reducing their reliance on China. This strategic move raises questions about the future dynamics of the semiconductor industry and the geopolitical implications for the...

#LLM On-Premise #DevOps
2026-02-11 DigiTimes

SMIC follow-up: dissecting the AI memory cycle and margin risk

An in-depth analysis of SMIC's role in the memory cycle for artificial intelligence applications, focusing on production capacity dynamics and potential risks to profit margins. Implications for the semiconductor market and the costs of implementing ...

#Hardware #LLM On-Premise #DevOps
2026-02-11 The Register AI

Open Compute taps IOWN to help design distributed datacenters

The Open Compute Project (OCP) aims to develop specs for distributed datacenters. The collaboration with IOWN (Innovative Optical and Wireless Network) seeks to leverage all-optical technologies to overcome the limitations of traditional connections,...

#LLM On-Premise #DevOps
2026-02-11 DigiTimes

AI memory demand threatens industrial PC margins and delivery

The increasing demand for memory in AI applications is creating challenges for industrial PC manufacturers. Margins are shrinking and delivery times are lengthening due to component shortages and rising costs.

#LLM On-Premise #DevOps
2026-02-10 TechCrunch AI

Ice Dance Duo Skates to AI Music at the Olympics: Plagiarism Issues

Czech ice dancers Katerina Mrazkova and Daniel Mrazek discovered that large language models (LLMs) can generate musical pieces that, unexpectedly, turn out to be plagiarism. This experience raises questions about originality and copyright in the age ...

#LLM On-Premise #DevOps
2026-02-10 Phoronix

Linux 7.0: Mainline Support for SpacemiT K3 RVA23 and Qualcomm Kaanapali

Linux kernel 7.0 introduces support for the SpacemiT K3 RVA23 and Qualcomm Kaanapali SoCs. This integration marks a significant step forward for the adoption of ARM and RISC-V architectures, expanding the hardware options available to developers and ...

#Hardware #LLM On-Premise #DevOps
2026-02-10 Tom's Hardware

AI Boom: Memory Makers Set to Earn $551 Billion by 2026

The memory market is poised to benefit significantly from the growth of artificial intelligence. Forecasts indicate that revenues will reach $551 billion by 2026, driven by strong demand from data centers. This figure is twice the revenue expected fo...

#LLM On-Premise #DevOps
2026-02-10 Tech.eu

UK bets on AI: chipmaker Fractile invests £100 million

The UK government is urging tech startups to take bolder risks in the artificial intelligence sector, promising support and investment. Fractile, a company specializing in chips for LLM inference, will invest £100 million in the UK to expand its site...

#Hardware #LLM On-Premise #DevOps
2026-02-10 Tech.eu

xWatts closes £1.6M to expand AI-powered energy management solutions

London-based xWatts, an intelligent energy management platform focused on decarbonising complex real estate assets, has closed a £1.6 million seed funding round. The company develops AI- and machine-learning-based technology to manage energy use acro...

#LLM On-Premise #DevOps
2026-02-10 DigiTimes

Phison CEO meets India's Modi to boost NAND, edge AI strategies

Phison's CEO met with Indian Prime Minister Modi to discuss expansion strategies in the Indian market, focusing on NAND memory and edge AI solutions. The initiative aims to strengthen Phison's presence in a rapidly growing market.

#LLM On-Premise #DevOps
2026-02-09 DigiTimes

SK's US$10bn AI venture takes chairman to Nvidia's door

SK Group chairman Chey Tae-won met with Nvidia executives as the Korean group invests US$10 billion in AI-related ventures. The meeting underscores the importance of collaborating with leading hardware providers for the implementation of large-scale ...

#Hardware #LLM On-Premise #DevOps
2026-02-09 Tom's Hardware

Taiwan rejects transfer of semiconductor capacity to the U.S.

Taiwan has rejected the possibility of transferring 40% of its semiconductor production capacity to the United States. Production increases in Taiwan are expected to occur in lockstep with production increases in the U.S.

2026-02-09 Phoronix

AMD Linux Driver Readying Peak Tops Limiter "PTL" Support

AMD is implementing support for the Peak Tops Limiter (PTL) in the AMDGPU and AMDKFD Linux kernel graphics drivers. This feature, intended for Instinct accelerators, aims to manage and limit peak power consumption.

#Hardware #LLM On-Premise #DevOps
2026-02-09 DigiTimes

AI spending spree threatens big tech cash flows

The acceleration of investments in the artificial intelligence sector is putting pressure on the cash flows of major technology companies. The need to support the growing demand for computational resources for training and inference of increasingly c...

#Hardware
2026-02-08 DigiTimes

India's budget to boost AI and chip ecosystem: implications

India's annual budget is set to provide a significant boost to the artificial intelligence and semiconductor ecosystem. The initiative aims to position India as a global technology hub, with targeted investments in research and development, infrastru...

#LLM On-Premise #DevOps
2026-02-08 DigiTimes

AI boom drives Taiwan's fastest growth in 15 years

Taiwan's economic growth accelerates due to strong demand in the artificial intelligence sector, overcoming fears of hollowing-out. Increased demand for high-performance semiconductors, essential for AI workloads, is a key factor in this expansion.

#Fine-Tuning
2026-02-08 LocalLLaMA

Strix Halo Distributed Cluster: LLM Inference with RDMA RoCE v2

A two-node cluster based on AMD Strix Halo, interconnected via Intel E810 (RoCE v2), has been built for distributed LLM inference using Tensor Parallelism. Benchmarks and setup guide are available online, opening new possibilities for local model exe...

#Hardware #LLM On-Premise #DevOps
2026-01-29 TechCrunch AI

SpaceX and xAI in talks to merge, report says

Reportedly, discussions are underway for a potential merger between SpaceX and xAI, the artificial intelligence company founded by Elon Musk. The deal would bring together the Grok chatbot, Starlink satellites, and SpaceX rockets under a single corpo...

2026-01-29 Tom's Hardware

Nvidia CEO: Taiwan's chipmaking capacity is secure

Nvidia CEO Jensen Huang denies rumors of a 40% shift in Taiwan's chip production capacity to the U.S. Huang states that TSMC's global expansion is essential to meet the high demand for leading-edge semiconductors, and Taiwan will maintain its key rol...

#Hardware #LLM On-Premise #DevOps
2026-01-28 DigiTimes

Taiwan tops auto display revenue while China dominates volume

According to DIGITIMES, Taiwan leads in revenue in the automotive display market, while China holds the leadership in terms of production volumes. This reflects the different strategies and specializations of the two regions in the automotive sector.

2026-01-28 Phoronix

Mesa NVK Driver Temporarily Disabling Support For Larger Pages Due To Bug

The Mesa NVK open-source Vulkan driver has temporarily disabled support for larger memory pages due to a discovered issue. This feature, implemented in the Nouveau kernel driver in Linux 6.19, aimed to improve performance through compression and the ...

#LLM On-Premise #DevOps
2026-01-28 TechCrunch AI

Waabi raises $1B and expands into robotaxis with Uber

Waabi has secured a $750 million Series C round to accelerate autonomous trucking. Additionally, Uber has committed roughly $250 million to deploy 25,000 robotaxis exclusively on its platform.

#LLM On-Premise #DevOps
2026-01-28 DigiTimes

ASML executives bet on chip demand strength past 2025

ASML executives predict strong chip demand beyond 2025. The company, a leader in the production of semiconductor lithography systems, expects continued growth in the sector, despite short-term market fluctuations.

2026-01-28 ArXiv cs.LG

Latent Structural Similarity Networks for Multivariate Time Series

A novel approach for analyzing multivariate time series using latent structural similarity networks. The method employs an unsupervised sequence-to-sequence autoencoder to learn window-level representations, aggregates these representations into enti...

#Fine-Tuning
2026-01-28 ArXiv cs.LG

NavFormer: IGRF Forecasting in Moving Coordinate Frames

NavFormer is a novel approach for forecasting the International Geomagnetic Reference Field (IGRF) in moving coordinate frames. It uses rotation invariant scalar features and a Canonical SPD module to stabilize the spectrum of window level second mom...

2026-01-28 LocalLLaMA

Stanford Study: Parallel Coding Agents, a Scam?

A study by Stanford and SAP questions the effectiveness of parallel coding agents. The findings indicate that adding a second agent significantly reduces performance due to coordination and communication issues. This raises doubts about platforms pro...

#LLM On-Premise #DevOps
2026-01-28 DigiTimes

Nexperia crisis: impacts on 'non-red' automotive supply chains

The crisis involving Nexperia, a semiconductor manufacturer, continues to generate significant repercussions on automotive supply chains, particularly for companies outside the 'red' sphere of influence. The implications extend to various levels of p...

#LLM On-Premise #DevOps
2026-01-28 DigiTimes

Analysis: Tesla splitting AI5 chip production between TSMC and Samsung

Tesla has decided to diversify the production of its AI5 chip, intended for artificial intelligence applications, relying on both TSMC and Samsung. This strategic move may be motivated by the desire to mitigate supply chain risks and optimize product...

#LLM On-Premise #DevOps
2026-01-28 DigiTimes

Jensen Huang prioritizes "empty chairs" over wrong hires at Nvidia

Nvidia CEO Jensen Huang emphasizes the importance of avoiding unsuitable hires, preferring to keep positions open rather than compromising team quality. Nvidia reportedly has a "bench" of approximately 60 potential future CEOs.

#Hardware #LLM On-Premise #DevOps
2026-01-27 Phoronix

AMD Radeon Linux Driver Introduces Low-Latency Video Decode Option

AMD's RadeonSI Gallium3D driver is introducing a new low-latency video decode mode in the upcoming Mesa 26.1 release. This feature offers faster response times but potentially increases GPU power consumption. A trade-off to consider for latency-sensi...

#Hardware #LLM On-Premise #DevOps
2026-01-26 DigiTimes

Analysis: Why Foxconn and Hotai are targeting Japan's toughest market

According to DIGITIMES, Foxconn and Hotai, two giants in the manufacturing and automotive sectors, are focusing their efforts on the Japanese market. This strategic move aims to capitalize on new opportunities in a notoriously complex and competitive...

#LLM On-Premise #DevOps
2026-01-26 DigiTimes

Samsung poised to start HBM4 shipments to Nvidia, AMD in February

Samsung is preparing to start shipments of its HBM4 memory to Nvidia and AMD starting in February. This new generation of high-bandwidth memory is crucial for accelerating the performance of GPUs and high-performance computing systems, essential for ...

#Hardware #LLM On-Premise #DevOps
2026-01-26 DigiTimes

Leading US defense companies deepen partnerships with Taiwan

Leading US defense companies are deepening their partnerships with Taiwan. This initiative aims to strengthen the island's technological and defense capabilities in an increasingly complex geopolitical context. The cooperation may include technology ...

#LLM On-Premise #DevOps
2026-01-26 DigiTimes

Memory innovation demand urgent as startups and SMEs gain opportunities

The increasing demand for memory capacity is driving innovation in the sector. Startups and SMEs face new opportunities to develop advanced solutions, essential to support artificial intelligence workloads and other data-intensive applications. Techn...

#LLM On-Premise #DevOps
2026-01-26 Tech.eu

2150 closes €210M Fund II for urban and industrial climate tech

Venture capital firm 2150 announced the final close of its €210 million second fund, bringing total assets under management to €500 million. The fund focuses on technology companies seeking to reshape cities and the industries that power them, with a...

2026-01-26 DigiTimes

Jabil names longtime director Steve Raymund chair

Jabil, a manufacturing services company, has announced the appointment of Steve Raymund as chair of the board of directors. Raymund, a current board member, succeeds Timothy Main.

2026-01-26 DigiTimes

Power transmission efficiency shapes future of AI computing

Power transmission efficiency is becoming a critical factor in the development of artificial intelligence and regional competitiveness. Optimizing energy consumption is essential to reduce the operating costs and environmental impact of AI data cente...

#Hardware #LLM On-Premise #DevOps
2026-01-25 Phoronix

LACT 0.8.4 Brings Improved Overclocking UI For GPUs On Linux

LACT, the third-party user interface for managing GPUs on Linux, updates to version 0.8.4. Particularly popular among AMD Radeon enthusiasts, LACT offers an alternative control panel for AMD and Intel graphics cards, in the absence of official soluti...

#Hardware
2026-01-25 The Next Web

EU: Single company structure for startups, ban on 'high-risk' tech

The European Union is accelerating innovation with "EU Inc," a unified legal structure for startups. Simultaneously, it aims to eliminate technology suppliers deemed "high-risk" from critical infrastructure. These measures aim to strengthen the conti...

2026-01-25 Tom's Hardware

Open-source Intel 486 motherboard built from scratch in six months

An enthusiast has showcased an open-source Intel 486 motherboard that they claim was 'made from scratch' in under six months. The initial goal was Linux and Doom compatibility, but the project appears to offer much more. The M8SBC-486 motherboard rep...

#Hardware
2026-01-25 LocalLLaMA

What happened to moondream3? The state of the visual model

The Moondream3 visual model, unveiled last year, seems to have disappeared. Despite an MLX version being available, Llama.cpp implementations and public updates are missing. The community is wondering about the future of this promising project.

#LLM On-Premise
2026-01-25 Phoronix

Focusrite Forte USB Audio Interface To Be Supported By Linux 7.0

The Focusrite Forte 2-in, 4-out USB audio interface, a portable audio recording solution, will be supported by the mainline Linux 7.0 kernel. The patches are queued in the Linux kernel's sound subsystem development tree. While a convenient little dev...

#Hardware
2026-01-25 DigiTimes

Fan motor demand drives growth for Taiwan analog chip makers

Taiwanese analog chip makers are benefiting from strong demand in the fan motor sector. Growth is driven by the need for efficient cooling solutions in various sectors, from consumer electronics to data centers. This positive trend highlights Taiwan'...

← Back to All Topics