ai-hardware

Apple's new M6 and M5 Ultra chips ship alongside Core AI, a new framework for building and deploying AI models on Apple silicon, with M5 Ultra supporting 512GB of unified memory.

Cerebras unveiled its CS-4 rack-scale inference system, claiming 30x faster performance than GPUs, though independent analysis finds the chip inside is an overclocked WSE-3, not a new design.

South Korean startup XCENA raised a $135M Series B at a $570M valuation for its MX1 CXL memory device, betting AI's real bottleneck is memory, not compute.

At Computex 2026, Samsung showed its first HBM5 mockup with a 2nm base die and Heat Path Block cooling, with mass production not expected before 2028.

London-based Fractile closes a $220M Series B led by Accel, Founders Fund, and Factorial Funds to tape out SRAM-based inference chips promising radically faster and cheaper AI output generation.

AI chipmaker Cerebras priced its IPO $25 above the upgraded range on May 13, opened at $350 on Nasdaq the next morning, and closed its first session at $311.07 before slipping roughly 10% on Friday.

TSMC debuted its A13 and A12 process nodes and updated its full roadmap through 2029 at its North America Technology Symposium, while confirming A16 volume production slides to 2027.

AI chipmaker Cerebras submitted its S-1 for a CBRS listing on April 17, backed by a $20 billion multi-year OpenAI deal and a swing to profitability, after withdrawing its first IPO attempt over a national security review.

Alibaba's Damo Academy has announced the XuanTie C950, a 5nm RISC-V CPU core capable of natively running hundred-billion-parameter language models, marking RISC-V's first serious entry into high-end AI computing.

Jim Keller's Tenstorrent unveils a $9,999 liquid-cooled workstation with four Blackhole ASICs, 480 Tensix cores, and a fully open-source software stack, shipping globally in Q2 2026.

Tesla plans to formally launch its Terafab semiconductor facility on March 21, aiming to produce up to 200 billion custom AI chips annually at 2nm process technology.

OpenAI's smaller, speed-optimized Codex-Spark model runs on Cerebras' Wafer Scale Engine 3, marking the AI giant's first move off Nvidia silicon for production inference.