NVIDIA A100 (Product): news timeline & CVEs

US arrests tech CEO accused of smuggling $300M in Nvidia chips into China

The US arrested Earthmade CEO Greg Lui for allegedly smuggling over $300 million in Nvidia AI chips to China.

The US Department of Justice arrested Greg Lui, CEO of Earthmade Computer, accusing him of conspiring to smuggle servers containing export-controlled Nvidia A100 and H100 GPUs into China. The indictment alleges shipments worth more than $300 million were disguised with false paperwork and routed through Malaysia, Singapore, and Hong Kong from October 2023 through August 2026. Lui is charged under the Export Control Reform Act and with smuggling and money laundering, facing up to 20 years if convicted. Nvidia said alleged diversion is less than half of one percent of its products.

Ars Technica · AI18:39 UTC · 3h agoAI policy 4 sources

Californian accused of shipping $300M worth of Nvidia chips to China without Uncle Sam’s approval

Greg Lui was charged with allegedly smuggling roughly $300 million in Nvidia AI GPUs to China.

Federal prosecutors charged Greg Lui, a 38-year-old California business owner, with allegedly exporting about $300 million worth of Nvidia AI hardware to China without required licenses. Using Earthmade Computer Inc., he allegedly ordered servers with A100, H100, PNY GeForce RTX 4090, and GeForce RTX 5090 GPUs, routed them through Malaysia and Singapore, and had them forwarded to China between about October 2023 and August 12, 2026. Prosecutors say he received more than $176 million from two Malaysian transshipment companies. He faces Export Control Reform Act, Export Administration Regulations, outbound smuggling, and money-laundering counts carrying up to 50 years.

The Register · Securityupdated 18:39 UTC · 3h agofirst 15:20 UTC · 6h agoPolicy & legal 4 sources

Scalability Analysis of Distributed Kolmogorov-Arnold Network Training on High-Performance Computing Systems

An empirical study shows distributed Kolmogorov-Arnold Network training reaches 74.7% parallel efficiency at 8 A100 GPUs, with overheads driven by All-Reduce choices.

The study evaluates data-parallel Kolmogorov-Arnold Network (KAN) training on the FinisTerrae III supercomputer using up to 8 NVIDIA A100 GPUs across 4 nodes with PyTorch Distributed Data Parallel. Strong scaling yields 5.97x speedup and 74.7% parallel efficiency at 8 GPUs, comparable to conventional deep learning workloads, while communication overhead ranges from 1.3% to 6.1%, driven mainly by All-Reduce algorithm selection and inter-node latency rather than KAN's edge-wise gradient structure. Weak scaling shows an initial single-to-multi-GPU throughput drop followed by stability, and the parameter-to-memory ratio improves with model size even as training time scales unfavorably. The authors provide GPU topology and model-size deployment guidelines for KAN training.

arXiv cs.AI / cs.LG / cs.CLSep 7, 16:45 UTC · 25d agoAI research1

NVIDIA A100

Timeline3 articles

Appears with

US arrests tech CEO accused of smuggling $300M in Nvidia chips into China

Californian accused of shipping $300M worth of Nvidia chips to China without Uncle Sam’s approval

Scalability Analysis of Distributed Kolmogorov-Arnold Network Training on High-Performance Computing Systems