AI Chips news and analysis on Web Pulse: 6197 curated articles tracking the latest AI Chips developments, tools, and research, updated continuously from vetted sources.
Meta Platforms Inc. is investing $50 billion in a new data center, raising questions about whether the AI boom is overbuilt, yet Wall Street responded positively to the announcement. The massive expenditure underscores M…
Nvidia on July 1 announced software optimizations that boost token throughput by up to 5x for DeepSeek V4 on Blackwell systems, with throughput improvements reaching 20x compared to baseline configurations on the same ch…
TSMC is building two new advanced packaging plants at Chiayi Science Park in southern Taiwan, bringing the total to four facilities that are projected to generate over 300 billion Taiwan dollars ($9.35 billion) in annual…
Global smartphone shipments fell 11% in Q2 2026, hitting the lowest level since 2013, according to Counterpoint Research, as a memory chip shortage driven by AI demand drives up prices. Samsung and Apple saw resilient de…
SK Hynix suffered its biggest one-day share price fall on record, sliding more than 15% in Seoul and wiping billions from its market value, just one trading session after its Nasdaq debut. The decline, triggered by a Kor…
Apple is developing the M7 Ultra chip with a potential 1.5TB unified memory capacity, targeting a 2028 release and aiming to deliver AI performance comparable to Nvidia's Blackwell accelerators, according to Mark Gurman'…
SK Hynix shares fell over 9% on the second day of US trading after the South Korean memory chip giant raised $26.5 billion in the largest-ever US listing by a foreign company, pricing 177.9 million ADRs at $149 each on N…
Meta disclosed Sunday that its Hyperion AI data center supercluster in Richland Parish, Louisiana, will expand to 5 gigawatts of compute capacity, pushing total project investment past $50 billion. The commitment makes H…
Infineon Technologies and South Korea's LS Electric signed a non-binding memorandum of understanding on July 10 to co-develop direct current power infrastructure for AI data centers, targeting power conversion systems, s…
DeepInfra, a Palo Alto-based GPU inference platform, opened a 1.7MW data center in Toronto on July 8, housing more than 1,000 NVIDIA Blackwell B300 GPUs in its first deployment outside the United States. The facility, bu…
Discarded NVIDIA enterprise Tesla GPUs, including K80 (24 GB, ~$60), P100 (16 GB, ~$75), and V100 (16 GB, under $200), remain usable for local AI inference and training workloads when paired with older CUDA toolkits and …
Apple is accelerating its chip release roadmap, skipping high-end M6 variants to reach the M7 generation faster, driven by AI competition from Nvidia and others, according to Bloomberg's Mark Gurman. The M7 Ultra process…
SK Hynix shares plunged more than 15% in Seoul on Monday, the stock's worst single-day decline on record, just three sessions after its $26.5 billion Nasdaq ADR listing. The crash triggered a 20-minute trading halt on th…
ASRock Rack unveiled the 2UXGI-THOR, a 2U short-depth edge server based on NVIDIA's IGX Thor industrial SoC, at Computex 2026. The server combines 14 Arm Neoverse-V3AE CPU cores with an NVIDIA Blackwell integrated GPU an…
Microsoft published a 109-page technical report on its MAI-Thinking-1 reasoning model, detailing the entire training process from web scraping to final optimization. The report reveals that 54.6% of the final training mi…
AI chip expectations are impossible to satisfy, said Gina Martin Adams, chief equity strategist at HB Wealth, as the sector's 83% surge earlier this year gives way to volatility and a reality check. The pullback reflects…
Intel is investing 5 billion euros ($5.7 billion) to expand its Leixlip campus in Ireland, focusing on Fab 34, one of the few EUV-capable facilities in Europe. The spending, roughly 30% of Intel's 2026 capital expenditur…
Meta and Amazon are leading a $725 billion capital expenditure surge among Big Tech companies for 2026, up from $410 billion in 2025, driven by AI infrastructure costs. Morgan Stanley projects total U.S. hyperscaler spen…
Dongfang Suanxin, a Shanghai chip startup founded in 2024 and led by Tsinghua professor Wei Shaojun, emerged from stealth with the DF1000 series, a 14nm AI processor delivering 520 teraflops of BF16 performance that uses…
Valeo System Architecture and Strategy Manager Frank Moesle told the 2026 Embedded Vision Summit that the automotive industry is transitioning from decentralized ECUs to zonal and central compute architectures, reducing …