DeepSeek launched a free, open-source large language model in late December, claiming it was developed in just two months at a cost of under $6 million — a much smaller expense than the one called for by Western counterparts. Last week, the company released a reasoning model that also reportedly outperformed OpenAI’s latest in many third-party tests.
“DeepSeek is an excellent AI advancement and a perfect example of Test Time Scaling,” an Nvidia spokesperson said. “DeepSeek’s work illustrates how new models can be created using that technique, leveraging widely-available models and compute that is fully export control compliant. Inference requires significant numbers of NVIDIA GPUs and high-performance networking. We now have three scaling laws: pre-training and post-training, which continue, and new test-time scaling.”
In a social media post, Marc Andreesen called DeepSeek’s product “one of the most amazing and impressive breakthroughs I’ve ever seen” and a “profound gift to the world.” The Andreessen Horowitz co-founder recently gained notoriety for his support of President Donald Trump.
These developments have stoked concerns about the amount of money big tech companies have been investing in AI models and data centers, and raised alarm that the U.S. is not leading the sector as much as previously believed.
“DeepSeek clearly doesn’t have access to as much compute as U.S. hyperscalers and somehow managed to develop a model that appears highly competitive,” said Srini Pajjuri, semiconductor analyst at Raymond James, in a Monday note.
.
.
>DeepSeek launched a free, open-source large language model in late December, claiming it was developed in just two months at a cost of under $6 million — a much smaller expense than the one called for by Western counterparts. Last week, the company released a reasoning model that also reportedly outperformed OpenAI’s latest in many third-party tests.
>“DeepSeek is an excellent AI advancement and a perfect example of Test Time Scaling,” an Nvidia spokesperson said. “DeepSeek’s work illustrates how new models can be created using that technique, leveraging widely-available models and compute that is fully export control compliant. Inference requires significant numbers of NVIDIA GPUs and high-performance networking. We now have three scaling laws: pre-training and post-training, which continue, and new test-time scaling.”
>In a social media post, Marc Andreesen called DeepSeek’s product “one of the most amazing and impressive breakthroughs I’ve ever seen” and a “profound gift to the world.” The Andreessen Horowitz co-founder recently gained notoriety for his support of President Donald Trump.
>These developments have stoked concerns about the amount of money big tech companies have been investing in AI models and data centers, and raised alarm that the U.S. is not leading the sector as much as previously believed.
>“DeepSeek clearly doesn’t have access to as much compute as U.S. hyperscalers and somehow managed to develop a model that appears highly competitive,” said Srini Pajjuri, semiconductor analyst at Raymond James, in a Monday note.
.
.
[Source](https://www.cnbc.com/2025/01/27/nvidia-falls-10percent-in-premarket-trading-as-chinas-deepseek-triggers-global-tech-sell-off.html)
(post is archived)