Justice Department probes include Nvidia, Megaspeed and Super Micro Computer

Justice Department probes include Nvidia, Megaspeed and Super Micro Computer

Live•Nov 13, 08:30 PM•2 min read•
Sarah Chen
Sarah Chen

Bloomberg reported, based on an anonymous source, that the Justice Department is probing Nvidia, including two focused on Megaspeed and Super Micro Computer, first reported by Fortune. A Manhattan inquiry examines whether Alice Huang-linked entities were involved in alleged efforts to smuggle Supermicro hardware laced with Nvidia GPUs to China; neither Nvidia nor Supermicro has been formally charged with wrongdoing.

One Justice Department inquiry concerns antitrust issues; two others focus on Megaspeed, where Alice Huang was a former top executive, and Super Micro Computer. The Manhattan inquiry’s connection to smuggling cases already brought in the United States or Asia remains unclear.

Separately, federal prosecutors charged Supermicro co-founder Yih-Shyan “Wally” Liaw and two others in March with allegedly routing about $2.5 billion in servers carrying Nvidia H200 and B200 GPUs to China through a Southeast Asian front company. Supermicro was not charged in that case and said it was a victim of the scheme; Liaw has pleaded not guilty.

Supermicro is an important Nvidia partner, and its CEO Charles Liang and Jensen Huang have been friends for three decades. Singapore’s Ministry of Home Affairs said the Singapore Police Force and Customs opened investigations into Megaspeed and Alice Huang and sought information from Supermicro. The ministry said the company offered “limited cooperation”; Singapore officials raised the matter with U.S. authorities in July 2025 and repeated requests to the Justice Department and FBI.

Singapore authorities sought information from Supermicro, but no public evidence so far directly connects the Megaspeed and Supermicro cases. Supermicro said in August that its board investigation found no evidence current senior management knew of the alleged scheme; it reported an unspecified number of terminations in sales, business development and technical support for policy or code-of-conduct failures.

Supermicro’s compliance history includes a 2006 guilty plea to illegally exporting computer equipment to Iran through a distributor in Dubai. An SEC accounting probe ended with a $17.5 million settlement in 2020.

Nvidia is slated to report third-quarter results on November 17, with Kress in place of Huang. House lawmakers requested a closed-door grilling with Huang the following day.

Six Chinese AI models carry Blackwell initialization fingerprints

Six models from DeepSeek, Moonshot AI, MiniMax and Ant Group carry weight patterns matching NVIDIA Blackwell GPU configurations, first reported by Unemploymentlabs. The evidence points to Blackwell-compatible initialization for those models, but it does not establish where the chips were located, how access was obtained or whether export controls were violated.

The six models are DeepSeek-V4-Flash and V4-Pro, Moonshot AI’s Kimi K3 and K3.1-Pro, MiniMax-M3, and Ant Group’s Ling-3.1-flash. V4-Flash and MiniMax-M3 match the 148-SM B200/B300 layout; the other four match the 152-SM GB200/GB300 layout.

The fingerprints come from initialization patterns preserved in publicly released weights. Recovering the seed and GPU thread layout identifies the hardware that generated starting values, not necessarily every accelerator used later.

Earlier Hopper fingerprints appear in DeepSeek-V3, Kimi K2.5, MiniMax-M2.7 and Ling-3.0-flash. Kimi K2.7-Code is an intermediate case: its Hopper-era weights include a small part matching B200/B300. Blackwell-matching releases include DeepSeek-V4 in April 2026, MiniMax-M3 in June, Kimi K3 in July, Ling-3.1-flash in October and Kimi K3.1-Pro in November.

The release sequence shows a shift in initialization fingerprints across four labs, not proof that entire training runs used Blackwell.

The strongest evidence for the six models ranged from 49 to 1,385 sigma, while the weakest individual test reached 20 sigma. A specially configured AMD MI300X can reproduce the layouts; for V4-Flash, 363 of 377 values that differed between AMD and NVIDIA matched NVIDIA, and none matched AMD. The same test could not settle the other five models, though an AMD explanation is considered very unlikely.

Run estimates for five models put each at roughly 500 to 3,000 Blackwell GPUs over two to three months, assuming the hardware that initialized a model also trained it. At roughly $70,000 per GPU, the largest cluster would cost $150 million to $200 million, and the five final runs together more than half a billion dollars; total lab spending on export-controlled GPUs is likely in the billions.

A U.S. license is required to export these Blackwell chips to China, and none has been reported.

The newest readable models from Alibaba (Qwen), Zhipu (GLM), Tencent, Xiaomi and StepFun still matched Hopper. Meituan says LongCat-2.0 was trained entirely on AI ASIC superpods, and its released weights showed no NVIDIA fingerprint.