Key Takeaways
- Cerebras Systems will provide Wafer-Scale Engine processors for AWS data center infrastructure targeting AI inference workloads.
- Partnership spans multiple years; neither company revealed financial details of the agreement.
- According to Cerebras, its technology delivers inference processing speeds up to 25 times faster compared to Nvidia’s GPU offerings.
- In January 2026, OpenAI committed to a Cerebras deal exceeding $10 billion in value.
- The chip startup secured $1 billion in funding during February 2026, achieving a valuation near $23 billion.
Amazon Web Services has entered into a long-term collaboration with emerging semiconductor company Cerebras Systems, planning to integrate its Wafer-Scale Engine technology throughout AWS infrastructure. The strategic focus centers on AI inference operations — the computational stage where artificial intelligence models generate responses to user prompts.
As the world’s premier cloud computing platform, AWS has traditionally relied on proprietary silicon solutions, particularly its Trainium processors engineered by subsidiary Annapurna Labs. This latest partnership will see AWS merging Trainium capabilities with Cerebras technology to deliver enhanced inference performance.
According to Cerebras, its Wafer-Scale Engine excels during the “decode” phase of inference operations — the critical moment when AI models formulate their outputs — achieving performance levels up to 25 times superior to Nvidia’s graphics processing units.
AWS will market this capability as a high-performance tier. “Those seeking slower inference speeds will find more economical alternatives available,” stated Cerebras CEO Andrew Feldman. AWS indicated it would continue providing budget-friendly inference options powered exclusively by Trainium processors.
Major Milestone for Cerebras Systems
This AWS arrangement follows closely on the heels of another landmark agreement OpenAI finalized with Cerebras during January 2026, with reports suggesting the contract exceeds $10 billion. That partnership aims to fuel OpenAI’s ChatGPT platform using Cerebras silicon, with OpenAI targeting the installation of computing infrastructure capable of 750 megawatts.
During February 2026, Cerebras completed a $1 billion capital raise, pushing its cumulative funding to $2.6 billion while establishing a company valuation around $23 billion. Investment participants include Fidelity Management, Benchmark, Tiger Global, and Coatue.
Cerebras initiated public offering preparations in September 2024 but subsequently withdrew those plans approximately twelve months later.
Mounting Competition for Nvidia
The collaboration between AWS and Cerebras represents another competitive hurdle for Nvidia within the inference computing segment. Industry momentum has been transitioning from model training activities, where Nvidia’s GPUs maintain clear leadership, toward inference applications requiring superior processing velocity.
Nvidia remains proactive in response. During December 2025, the company finalized a $20 billion licensing arrangement with chip manufacturer Groq. Additionally, Nvidia has announced plans to introduce innovative processing architecture incorporating Groq’s innovations in the immediate future.
Nafea Bshara, co-founder of Annapurna Labs at AWS, emphasized that the Cerebras collaboration prioritizes accelerating performance while reducing expenses. “We’re committed to maximizing speed and minimizing cost,” he explained.
Shares of Amazon’s AMZN stock declined 0.44% at the time of this report.
