Supermicro Q1: Sales $19.9B, Margin Falls to 12.7%
Supermicro Q1 FY2027 sales were $19.9B with a 12.7% gross margin, down from 17.5%. Q2 guide is $19.5B to $21.5B; FY2027 rises to $72B to $82B.

Supermicro Q1 FY2027 sales were $19.9B with a 12.7% gross margin, down from 17.5%. Q2 guide is $19.5B to $21.5B; FY2027 rises to $72B to $82B.

Caiwen and TechNode report DeepSeek plans to train V5 on Huawei Ascend 950 chips. Neither report has a DeepSeek or Huawei statement.

AMD's Q3 2026 revenue rose 45% to $13.4 billion as data center sales jumped 97% to $8.5 billion. MI455X began volume shipments, Q4 guidance is about $17 billion, and AMD expects 2027 revenue growth above 80%.

There is no single US ban on AI chips for China. A plain-English guide to the export controls: how a chip is classified, the tools and memory rules, who needs a license, how the rules changed from 2022 to 2026 and what they mean for Chinese AI labs.

Sandisk's fiscal Q1 2027 revenue rose 378% to $11.0 billion as datacenter sales hit $4.8 billion. It guided Q2 to $12.2 billion to $12.6 billion.

Intel's Q3 revenue rose 21% to $16.5 billion as data center sales jumped 78%, but a PC slowdown left its Q4 forecast below analyst estimates.

AWS sales rose 39% to $46.0 billion in Q3 2026, its fastest growth in nearly five years. Amazon kept its 2026 capex plan at about $220 billion.

Iluvatar CoreX showed its Tiangai 300 GPU live in Shanghai and claims it matches Nvidia's H200. The claimed numbers, the IXL144 rack, and what is unverified.

China's first 1,024-card Huawei Ascend 950DT SuperPoD went live in Wuhan on October 28. The reported specs, the 950 roadmap, and what is still missing.

Nvidia's spec page now lists Vera Rubin GPUs at 19.2 TB/s of HBM4 and 3 TB/s of NVLink, down from 22 and 3.6 TB/s. Nvidia calls the old figures targets.

AI hyperscaler demand quintupled flash prices in a year, per TechInsights. Apple is expected to spend $25 billion on flash that cost $6 billion in 2025.

SK hynix posted record Q3 revenue of 104.6 trillion won and an 80% operating margin, and expects memory demand to exceed supply through 2030.

OpenAI plans to buy Google's TPU chips, The Information and Bloomberg report, citing unnamed sources. No terms are public, and neither company confirmed.

Samsung posted record Q3 operating profit of KRW 107.4 trillion. HBM4 revenue tripled, and Samsung says its 2027 HBM supply is already sold out.

Marvell is in advanced talks to buy optics maker Lumentum, valued at about $83 billion, Bloomberg and Reuters report. What each company does and what is open.

FuriosaAI raised a $620M Series D at a $2.9B valuation. Its RNGD inference card is in mass production, and a 2nm Gen 3 chip is in development with Broadcom.

Malaysian customs seized 30 servers with advanced AI chips at KLIA, valued at RM49.5 million (US$12.1 million). Officials call it the fifth such case this year.

A TPU is Google's custom chip for machine learning. Here is what it does, how it handles the matrix math inside AI models, how pods and slices work, how it compares with GPUs and CPUs, what the Edge TPU is, and who can use one, in plain English.

AMD will invest $1 billion in Meta's equity offering, and its first Helios racks are being tested at Meta. What runs today, what is planned, and the 6 GW deal.

Oracle says Google TPUs will come to OCI in 2027, the first hyperscale cloud outside Google to offer them, with Cerebras, Jalapeño and more.
