Cerebras becomes the World’s Fastest Host for DeepSeek R1, Outpacing Nvidia GPUs By 57x

Join our daily and weekly newsletters for the most current updates and exclusive material on industry-leading AI coverage. Discover more

Cerebras Systems revealed today it will host DeepSeek’s breakthrough R1 expert system model on U.S. servers, promising accelerate to 57 times faster than GPU-based options while keeping delicate data within American borders. The move comes in the middle of growing concerns about China’s fast AI advancement and information personal privacy.

The AI chip startup will deploy a 70-billion-parameter variation of DeepSeek-R1 running on its exclusive wafer-scale hardware, providing 1,600 tokens per second – a significant enhancement over conventional GPU implementations that have struggled with newer “thinking” AI models.

Why DeepSeek’s reasoning designs are reshaping business AI

” These thinking models affect the economy,” stated James Wang, a senior executive at Cerebras, in an exclusive interview with VentureBeat. “Any understanding worker generally needs to do some kind of multi-step cognitive tasks. And these thinking designs will be the tools that enter their workflow.”

The statement follows a turbulent week in which DeepSeek’s emergence activated Nvidia’s largest-ever market price loss, nearly $600 billion, raising questions about the chip giant’s AI supremacy. Cerebras’ service straight deals with 2 key concerns that have emerged: the computational needs of innovative AI models, and data sovereignty.

” If you utilize DeepSeek’s API, which is preferred right now, that data gets sent straight to China,” Wang described. “That is one extreme caveat that [makes] numerous U.S. business and business … not ready to think about [it]”

How Cerebras’ wafer-scale technology beats standard GPUs at AI speed

Cerebras attains its speed benefit through a novel chip architecture that keeps entire AI designs on a single wafer-sized processor, removing the memory traffic jams that pester GPU-based systems. The company declares its execution of DeepSeek-R1 matches or goes beyond the performance of OpenAI’s exclusive designs, while running completely on U.S. soil.

The development represents a considerable shift in the AI landscape. DeepSeek, founded by previous hedge fund executive Liang Wenfeng, surprised the market by attaining advanced AI reasoning abilities supposedly at just 1% of the cost of U.S. rivals. Cerebras’ hosting service now offers American companies a method to utilize these advances while preserving data control.

” It’s really a great story that the U.S. research study labs offered this present to the world. The Chinese took it and improved it, but it has restrictions since it runs in China, has some censorship issues, and now we’re taking it back and running it on U.S. information centers, without censorship, without information retention,” Wang said.

U.S. tech leadership deals with brand-new concerns as AI innovation goes global

The service will be available through a developer preview beginning today. While it will be at first totally free, Cerebras strategies to execute API access controls due to strong early demand.

The relocation comes as U.S. lawmakers face the ramifications of DeepSeek’s increase, which has actually exposed potential constraints in American trade limitations developed to keep technological advantages over China. The ability of Chinese companies to achieve development AI abilities regardless of chip export controls has actually triggered calls for brand-new regulatory methods.

Industry experts suggest this advancement could accelerate the shift away from GPU-dependent AI facilities. “Nvidia is no longer the leader in inference efficiency,” Wang kept in mind, indicating standards showing remarkable efficiency from various specialized AI chips. “These other AI chip companies are really faster than GPUs for running these most current models.”

The impact extends beyond technical metrics. As AI designs significantly include advanced thinking capabilities, their computational needs have increased. Cerebras argues its architecture is better matched for these emerging work, potentially reshaping the competitive landscape in enterprise AI release.

If you desire to impress your boss, VB Daily has you covered. We offer you the inside scoop on what companies are doing with generative AI, from regulatory shifts to useful implementations, so you can share insights for optimum ROI.

Read our Privacy Policy

An error occured.

The AI Impact Tour Dates

Join leaders in enterprise AI for networking, insights, and interesting discussions at the upcoming stops of our AI Impact Tour. See if we’re pertaining to your area!

– VentureBeat Homepage
– Follow us on Facebook
– Follow us on X.
– Follow us on LinkedIn.
– Follow us on RSS

– Press Releases.
Us.
– Advertise.
– Share a News Tip.
– Add to DataDecisionMakers

– Privacy Policy.
– Terms of Service.
– Do Not Sell My Personal Information

© 2025 VentureBeat. All rights reserved.

AI Weekly

Your weekly take a look at how applied AI is altering the tech world

We appreciate your personal privacy. Your email will only be utilized for sending our newsletter. You can unsubscribe at any time. Read our Privacy Policy.

Thanks for subscribing. Check out more VB newsletters here.