📁 Hardware

China's next-generation Loongson 3B6600 (CPU) and 9A1000 (GPU) chips aim to match the performance of Intel 12th Gen and AMD RX 550 by 2027. This development highlights China's ambition to strengthen its technological autonomy in the semiconductor sector, with implications for data sovereignty and on-premise deployment strategies.

Applied Materials and TSMC have announced a collaboration at the EPIC Center to accelerate the development of chips dedicated to artificial intelligence. This initiative aims to optimize manufacturing processes and foundational technologies, with significant implications for the efficiency and availability of AI hardware, crucial for companies evaluating on-premise deployments and data sovereignty management.

VSO Electronics is targeting significant growth, driven by the increasing demand for specialized cables for AI infrastructure. The company also plans to activate a new in-house leak-detection line by late 2026, consolidating its production capabilities and quality control in a rapidly evolving market.

AcBel Polytech, OmniOn, and Kinpo Group have formed a strategic partnership to develop power supply solutions specifically for the growing artificial intelligence market. This initiative aims to address the demand for robust and efficient infrastructure, essential for intensive LLM workloads and on-premise deployments, where power efficiency and thermal management are critical factors for TCO.

A user demonstrated an open-loop tap water cooling method for a DGX system, keeping GPUs below 68°C at 95% utilization. The setup handles a Qwen3.5-122b-a10B LLM with Q6_K precision, utilizing 110 GB of memory and an 80k context window, achieving 18.77 tokens/second for continuous vision analyses. This highlights the challenges and creative solutions for on-premise AI deployments.

The open-source "R300g" driver for ATI (AMD) Radeon R300 and R500 series GPUs, dating back over two decades, is set to receive a significant code restructuring in 2026. This effort, led by a single community developer, highlights the longevity and dedication of open-source projects, ensuring support and improvements even for hardware considered obsolete.

Rumors suggest AMD is preparing an entry-level RDNA 4 GPU, the RX 9050, featuring 8GB of VRAM and 2048 cores. This potential addition to the Radeon lineup could offer new options for lighter AI workloads and on-premise deployments, balancing cost and capability for specific inference needs.

AMD has released significant updates for its AMDGPU kernel driver on Linux, introducing support for HDMI 2.1 Fixed Rate Link (FRL) and Display Stream Compression (DSC). These enhancements enable higher resolutions and refresh rates, solidifying the open-source driver's position as a robust solution for AMD hardware in environments demanding advanced graphics performance and infrastructural control.

System76 has unveiled the Thelio Major workstation, a high-end Linux system built entirely on AMD hardware. Featuring AMD Ryzen Threadripper 9000 series processors and Radeon AI PRO R9700 graphics, this machine offers a powerful, open-source solution ideal for developers and professionals requiring high performance for intensive workloads, including those related to artificial intelligence. It provides complete control over the operating environment and data sovereignty.

The technology industry, particularly in the field of artificial intelligence, is evolving at an unprecedented pace. For CTOs and infrastructure architects, keeping up means understanding the implications of new hardware developments and deployment strategies. This requires an in-depth analysis of on-premise options, costs, and data sovereignty, all crucial aspects for informed decisions.

Intel and SK Hynix shares surged following reports of a potential strategic chip packaging partnership. The collaboration would involve SK Hynix testing Intel's 2.5D EMIB technology for High Bandwidth Memory (HBM) integration. This move highlights the increasing importance of advanced packaging technologies for AI and LLM applications, with significant implications for performance and efficiency in next-generation hardware.

The upcoming Linux kernel version 7.2 will integrate new power management control features for AMD Ryzen AI and Intel NPU drivers. These optimizations, part of the `drm-misc-next` pull request, aim to improve efficiency and performance for AI workloads on local hardware, offering IT professionals greater control over on-premise deployments and contributing to better TCO analysis.

Dutch company eyeo has secured €40 million in a Series A funding round, bringing its total capital to €55 million. The funds will be used for the commercialization of its NCOS color-splitting image sensor technology, in-house chip design, and volume production. The goal is to accelerate the market adoption of this innovation, with significant implications for data acquisition in AI contexts.

German researchers have developed an innovative portable 40mm launcher designed to neutralize drones. This low-tech system employs a mechanical "bola," firing 6.5-feet-long steel chains at 80 meters per second. The approach stands out for its effectiveness against quadcopters, offering a mechanical alternative to more complex solutions like lasers or EMPs, and outperforming textile-based systems.

Dutch nanophotonic imaging company eyeo secured €40 million in a Series A funding round, bringing its total capital to €55 million. The startup develops nanophotonic technology for image sensors, enhancing light sensitivity, color accuracy, and resolution by replacing traditional color filters. The funds will support commercial expansion and the development of next-generation 3D-stacked CMOS sensors, with critical applications for Edge AI and autonomous systems.

LaceLocker® proposes a vision for the next generation of wearables, focusing on integrating connectivity into everyday objects, such as footwear. This approach aims for integrated hardware platforms that fit naturally into people's lives, fostering collaboration across technology sectors and moving beyond reliance on bulky devices.

A Micron executive highlights how memory limitations are an increasing challenge for GPU efficiency in data centers, especially with the escalation of AI inference workloads. This constraint directly impacts the scalability and TCO of deployments, requiring targeted hardware and software strategies to optimize performance and the management of large models.

The explosion of artificial intelligence inference workloads is fueling a "memory race" among leading manufacturers. Samsung is at the forefront of this competition, developing solutions that address the growing demand for VRAM and bandwidth. This dynamic has direct implications for companies evaluating self-hosted LLM deployments, impacting TCO and data management capabilities.

OpenAI and leading chip manufacturers are collaborating on a new initiative, dubbed MRC, aimed at mitigating critical slowdowns affecting artificial intelligence model training processes. This strategic move underscores the importance of optimizing both hardware and software infrastructure to support the development of increasingly complex LLMs, with significant implications for on-premise deployments.

Recent advancements demonstrate how the DeepSeek-V4-Flash model, optimized with MTP self-speculation and advanced quantization techniques, can achieve significant performance on on-premise hardware. Utilizing two NVIDIA RTX PRO 6000 Max-Q GPUs, each with 96 GB of VRAM, up to 85.52 tokens/second were recorded with a 524k token context, highlighting the potential for efficient LLM deployments in local environments.