DeepSeek Funding Locks In at Least RMB 80 Billion; Nvidia Market Cap Nears $6 Trillion

AI Capital and Compute Race Accelerates

On October 6, 2026, several AI-industry-related developments were released in quick succession, with the focus on financing, compute, and the commercialization progress of model companies.

DeepSeek Funding Approaches RMB 100 Billion

Citing people familiar with the matter, Bloomberg reported that Chinese AI company DeepSeek's latest funding round is nearing completion, having locked in at least RMB 80 billion (about USD 12 billion), far exceeding its initial funding target of about RMB 50 billion. According to signed term sheets, the final funding size could approach RMB 100 billion ^[3]. Tencent Holdings and CATL are the investors that have committed the largest amounts. DeepSeek initially sought only about RMB 50 billion in this round, but after the smooth release of its latest model, investor demand exceeded expectations, and the company's initial target valuation was about RMB 500 billion ^[3]. The report also mentioned that DeepSeek is building a data center in Inner Mongolia and deploying Huawei AI chips ^[3].

Nvidia Market Cap Hits Record High

On the same day, about twenty minutes after the Nasdaq opened, Nvidia (NVDA) reached a market capitalization of $5.82 trillion, a record high, just one step away from $6 trillion ^[3]. At the open, Nvidia's stock price was $242.10, with a high of $243.37 and a low of $240.76. At the close on October 5, U.S. Eastern Time, Nvidia closed at $238.90, up 2.12% for the day, marking its third consecutive trading day of record closing highs. If it continues to rise, Nvidia could become the world's first company with a market cap of $6 trillion ^[3].

Moonshot AI Funding Report and Big Tech Attitudes

In addition, reports said Moonshot AI had completed a $50 billion Pre-IPO funding round ^[3]. At the same time, Microsoft and Meta were reported to have asked employees to reduce their use of Claude ^[3], indicating that large tech companies' attitudes toward internal AI tool choices are changing.

OpenAI Math Papers and Academic Discussion

On model capabilities, OpenAI released 722 mathematics papers overnight, covering areas such as Riemann, Hodge, and BSD; mathematicians said they could not keep up, while three Fields Medal winners pointed out that this does not represent endorsement ^[4]. The first-day release of OpenAI's 'Crazy 28 Days' event also drew attention ^[2]. Around AI's role in mathematics and other basic research, Terence Tao was discussed as being in the AI slowdown camp ^[6].