42 articles tagged with #ai-hardware. AI-curated summaries with sentiment analysis and key takeaways from 50+ sources.
The U.S. government is reportedly considering sweeping new chip export controls that would give it oversight over every chip export sale globally, regardless of the originating country. This drafted proposal represents a significant expansion of U.S. regulatory reach in the semiconductor industry.
Researchers introduce Deep Optimizer States, a technique that reduces GPU memory constraints during large language model training by dynamically offloading optimizer state between host and GPU memory during computation cycles. The method achieves 2.5× faster iterations compared to existing approaches by better managing the memory fluctuations inherent in transformer training pipelines.
Broadcom's stock gains momentum following UBS's upgrade of AI revenue projections to $145 billion, driven by Google's extension of its TPU chip partnership through 2031 and increased compute allocation to Anthropic. The extended partnership signals sustained demand for specialized AI infrastructure and validates Broadcom's positioning as a critical supplier in the competitive AI hardware ecosystem.
Obex, a Sky-backed stablecoin incubator, is deploying $1 billion in USDS across eight projects focused on mortgages, AI hardware, and solar energy. This represents the protocol's largest effort to diversify yield sources beyond traditional crypto-native investments.
Sky-backed stablecoin incubator Obex is deploying $1 billion across tokenized assets in AI hardware, energy, and housing sectors to diversify yield sources beyond traditional crypto markets. This strategic move aims to expand Sky's ecosystem by targeting real-world assets rather than relying on circular cryptocurrency yields.
Nvidia stock rises following three major developments: Arm's new AI CPU launch, a massive $50B+ AWS order for 1 million GPUs, and resumed chip sales to China potentially worth $32B annually. These combined deals represent approximately $82B in new revenue streams for the semiconductor giant.
Nvidia announced the Groq 3 LPU at GTC 2024, its first chip specifically designed for AI inference rather than training, incorporating technology licensed from startup Groq for $20 billion. The chip uses SRAM memory integrated within the processor to achieve 7x faster memory bandwidth than traditional GPUs, optimizing for the low latency required for real-time AI inference applications.
Meta has unveiled four custom AI chips developed in partnership with Broadcom, claiming some outperform existing commercial silicon solutions. This move represents Meta's strategic shift toward developing proprietary AI hardware to reduce dependence on third-party chip manufacturers.
A research paper presents a 10-year roadmap for coordinated AI and hardware co-development, targeting 1000x efficiency improvements in AI training and inference by 2035. The vision emphasizes energy efficiency over raw compute scaling, proposing integrated solutions across algorithms, architectures, and systems to enable sustainable AI deployment from cloud to edge environments.
Meta faces a lawsuit over privacy concerns regarding its AI smart glasses, with allegations that the company's marketing promised user control while subcontractors were actually reviewing customer footage including sensitive content. The legal action centers on discrepancies between Meta's privacy promises and actual data handling practices.
The memory chip shortage is driven by massive AI demand for high-bandwidth memory (HBM), causing DRAM prices to surge 80-90% this quarter. While major AI companies have secured supply through 2028, other industries face scarce supply and inflated prices that won't normalize for years.
Researchers at UC San Diego developed a new type of bulk resistive RAM (RRAM) that overcomes traditional limitations by switching entire layers rather than forming filaments. The technology achieved 90% accuracy in AI learning tasks and could enable more efficient edge computing by allowing computation within memory itself.
AI data centers are hitting physical limits with copper cables as GPU-to-GPU data rates approach terabit-per-second speeds, requiring thicker, shorter cables that complicate dense connections. Startups Point2 Technology and AttoTude are developing radio-based cable solutions that promise longer reach, lower power consumption, and narrower cables than copper alternatives.
Researchers have developed a new approach to improve microelectronics energy efficiency by stacking multiple active components made from new materials on the back end of computer chips. This innovation aims to reduce energy waste during computational processes.
OpenAI and Foxconn have announced a partnership to design and manufacture AI infrastructure hardware domestically in the U.S. The collaboration aims to develop multiple generations of data-center systems and strengthen domestic supply chains for advanced AI infrastructure.
OpenAI and Broadcom announced a multi-year strategic partnership to deploy 10 gigawatts of OpenAI-designed AI accelerators by 2029. The collaboration will focus on co-developing next-generation systems and Ethernet solutions for scalable, energy-efficient AI infrastructure.
Hugging Face has acquired Pollen Robotics to expand into the open-source robotics market, enabling the AI platform company to sell physical robots alongside its existing AI model ecosystem. This acquisition represents Hugging Face's strategic move to bridge software and hardware in the AI/robotics space.
Tesla stock rose 0.66% in premarket trading following Elon Musk's announcement that the AI5 chip has completed tape-out. The company plans to begin mass production in 2027 while simultaneously advancing development on the AI6 chip, signaling continued momentum in Tesla's AI hardware ambitions.
Major tech companies including Meta and Amazon are investing billions in AI hardware with a 3-year useful lifespan, creating a sustainability and capital efficiency problem. The article suggests that consumers and businesses using AI products may benefit more than the hardware manufacturers themselves, raising questions about the long-term viability of the current AI infrastructure spending model.
Bernstein analysts warn that Nvidia could severely impact Supermicro by reducing GPU supply access, despite Supermicro's success being tied to Nvidia's $4 trillion valuation. The dependency relationship gives Nvidia significant leverage to potentially devastate Supermicro's hardware business at any time.
Meta and EssilorLuxottica are preparing to launch two new Ray-Ban AI glasses models, according to recent FCC filings describing production units. The filings suggest an imminent launch, following a similar timeline to their second-generation Ray-Ban release in late 2023.
Meta has developed four new MTIA processors designed to power its AI and recommendation systems. This represents the tech giant's continued effort to build proprietary AI hardware while still investing billions in equipment from industry leaders like Nvidia.
Qualcomm and Arduino have launched the Ventuno Q single-board computer featuring a dual-brain architecture designed specifically for AI and robotics applications. The device combines AI processing power with real-time control capabilities, positioning it as a potential competitor to existing market leaders in the robotics computing space.
Tesla billionaire KoGuan has purchased 1 million Nvidia shares and plans to acquire another 1 million, demonstrating confidence in AI infrastructure investments. This significant investment highlights the growing institutional interest in AI hardware companies and could influence broader market dynamics in the tech sector.
Researchers evaluated HiFloat (HiF8 and HiF4) formats for low-bit inference on Ascend NPUs, finding them superior to integer formats for high-variance data and preventing accuracy collapse in 4-bit regimes. The study demonstrates HiFloat's compatibility with existing quantization frameworks and its potential for efficient large language model inference.
