Semiconductors
Accelerators, HBM and memory, foundry capacity, and how chip supply sets the pace of AI build-outs.
AMD pitches native FP64 on Instinct MI430X as Nvidia leans on emulation at SC26
AMD used the opening of Supercomputing 2026 (SC26) in Chicago this week to promote native double-precision (FP64) support on its newest data center accelerator, the Instinct…
Saudi Infra commits up to SAR 2.5 billion to HUMAIN, with grid built for 1GW
Saudi Arabia's National Infrastructure Fund, known as Infra, has agreed to provide HUMAIN with up to SAR 2.5 billion (US$667 million, £505.3 million) across two financing…
Jensen Huang steps down as NVIDIA chief executive, with Colette Kress named interim CEO
Jensen Huang, who co-founded NVIDIA, has resigned as the company's chief executive, Data Center Frontier reports. NVIDIA has appointed Colette Kress, its chief financial officer…
Samsung lays out 144 trillion won Chungcheong buildout spanning packaging, OLED and AI substrates
Samsung plans to spend a combined 144 trillion won across its affiliates to expand chip, display, substrate and battery manufacturing in Korea's Chungcheong region. The group…
Report alleges Megaspeed executive helped route $7.5 billion in Nvidia AI GPUs into China
A Megaspeed executive, Alice Huang, allegedly played a central role in bringing more than $7.5 billion worth of Nvidia AI GPUs into China, Tom's Hardware reported, citing an…
Samsung trims a chip thermal analysis from 18 hours to 41 seconds on NVIDIA GPUs
Samsung Electronics reported that one chip-scale thermal and mechanical analysis, which previously ran for about 18 hours on a 32-core CPU, now finishes in 41 seconds on a GPU…
EU commits €525 million to a platform for new European chip design companies
The European Commission has launched the European Chip Design Platform, an initiative under the EU Chips Act that is set to mobilize €525 million in investment aimed at building a…
Broadcom TPU compute shipments double to nearly 2 million H100-equivalents
Broadcom's shipments of Google-designed Tensor Processing Units (TPUs) roughly doubled in compute terms between its fiscal Q2 and Q3 2026, the period from May to early August…
AMD guides fourth-quarter revenue to $17 billion, topping analyst estimates
Advanced Micro Devices Inc. expects fourth-quarter revenue of $17 billion, plus or minus 3%. The chipmaker issued the guidance in a statement on Tuesday, and the figure came in…
Apple's manufacturing bill jumps $33 billion as memory prices climb
Apple's cost to build its products has risen by $33 billion over just a few months, driven by sharply higher memory prices, Tom's Hardware reported, citing Reuters. The company…
Quarterly AI chip compute additions grow about 2.5× a year, doubling every 9 months
The computing capacity of AI chips sold worldwide each quarter has been rising about 2.5× per year since 2024, an analysis by Epoch AI finds. That implies a doubling time of 9.1…
OpenAI chief Sam Altman meets about 50 Samsung executives in South Korea over memory supply
OpenAI CEO Sam Altman traveled to South Korea and sat down with about 50 senior Samsung executives as OpenAI looks to grow its AI and memory capacity, Tom's Hardware reported…
Intel moves some client CPU capacity to Xeon server chips as AI demand surges
Intel is shifting part of its manufacturing capacity away from consumer processors and toward data center CPUs. The company expects the AI industry's appetite for anything that…

Malaysia’s server-trade gap turns permit enforcement into a capacity-risk signal
Epoch AI found a $3.8 billion versus $0.6 billion mirror-trade gap that narrowed after Malaysia tightened transit rules. The pattern is consistent with diversion, but it does not…