Global GPU Computing News【20251101】

NVIDIA Advances AI Computing Hub in Europe: NVIDIA, in partnership with Deutsche Telekom, plans to invest €1 billion to build an AI data center in Munich, Germany, deploying approximately 10,000 GPUs. SAP, Europe's largest enterprise software company, will be a primary user. The plan is expected to be officially announced in Berlin this November.

NVIDIA Deploys Large-Scale GPUs in South Korea: Collaborating with the South Korean government, Samsung, SK Group, Hyundai Motor, and Naver Cloud, NVIDIA plans to deploy a total of 260,000 Blackwell architecture GPUs in South Korea to build a large-scale "AI factory." This initiative will significantly increase South Korea's total AI computing capacity from 65,000 GPUs to over 300,000.

Clarifai and Vultr Demonstrate Leading AI Inference Performance: At the NVIDIA GTC AI conference, Clarifai and Vultr jointly announced record-breaking AI inference performance on GPUs. Their solution achieved a processing speed of 544 tokens per second on the GPT-OSS-120B model, with a cost of just $0.16 per inference.

Google and Anthropic Secure Major Chip Deal: Google will provide AI startup Anthropic with up to 1 million of its self-developed TPU chips in a deal worth tens of billions of dollars, deepening their collaboration in the AI field.

NVIDIA H100 GPU Planned for Space Mission: AI company Crusoe, in partnership with Starcloud, plans to send NVIDIA H100 GPUs into space by November 2025, claiming the AI computing power provided will be "100 times more powerful" than any computing capacity previously available in space.

Intbology and Shenzhou Guangda Form Strategic Partnership: The two companies will collaborate comprehensively on high-end GPU computing assurance, AI industry application innovation, and the development of domestic computing power, aiming to build a more complete and reliable AI infrastructure services ecosystem.

Moore Threads' Full-Function GPU Featured at CNCC2025: At the 22nd China National Computer Conference, Moore Threads showcased its practice of building efficient "AI factories" using domestic full-function GPUs, providing a computing foundation for large model training and inference.

Domestic GPU Company MuXi IPO Approved on STAR Market: MuXi successfully passed the listing review on the Sci-Tech Innovation Board (STAR Market), planning to raise 3.904 billion yuan for the development of new general-purpose GPUs and the industrialization of AI inference chips, marking a key technological breakthrough in the domestic high-end chip sector.

Increased Domestic AI Computing Investment and Policy Support: According to reports, Alibaba announced plans to invest 380 billion yuan over the next three years in cloud computing and AI infrastructure, while Tencent plans to invest over 100 billion yuan. Concurrently, domestic support for AI computing development includes issuing "AI vouchers" and launching city-wide "millisecond-level computing access" initiatives.

Strong Performance of AI-Themed ETFs: As of October 27, 2025, the AI Artificial Intelligence ETF (512930) has accumulated a year-to-date increase of 72%, reflecting the capital market's continued optimism towards the AI computing sector.