Gold bars to be exempt from tariffs, White House clarifies
Investing.com-- NVIDIA Corporation (NASDAQ:NVDA) on Thursday said China’s DeepSeek R1 artificial intelligence model was now available as a preview on its enterprise software platform, and that developers could access its application programming interface soon.
DeepSeek R1 is available as an Nvidia Inference Microservice (NIM) preview on the company’s website, Nvidia said in a statement. NIM is a service that allows developers to deploy and use AI programs on their personal systems while using remote Nvidia hardware.
“The DeepSeek-R1 NIM microservice can deliver up to 3,872 tokens per second on a single NVIDIA HGX H200 system,” Nvidia said.
The chipmaker said API support for DeepSeek is expected to be “available soon” as a downloadable NIM service.
Nvidia’s adoption of DeepSeek comes less than a day after Microsoft Corporation (NASDAQ:MSFT) made the new R1 model available on its Azure cloud platform.
DeepSeek shot up in popularity over the past week as its R1 model appeared to be able to match and sometimes surpass the performance of much bigger rivals such as OpenAI’s Chat-GPT, while using older hardware and a fraction of the budget.
The AI’s release sparked a steep decline in Nvidia’s market valuation, wiping out around $600 billion in capital this week as investors questioned the need for rapid investment in AI infrastructure.
Companies could instead use efficient, leaner AI models, cutting down the necessity for more data center infrastructure, which has served as a key source of demand for Nvidia over the past two years.
The company makes the most advanced AI chips in the market, and saw a massive spike in valuation over the past two years on the prospect of increased demand from the AI boom.