Amazon (AMZN) Stock: AWS Partners With Cerebras to Challenge Nvidia's AI Chip Dominance

Key Highlights

Table of Contents

Cerebras’s Wafer-Scale Engine chips will be integrated into AWS data center infrastructure for AI inference operations.
Partnership extends multiple years; specific financial details remain undisclosed.
According to Cerebras, its processors deliver inference performance up to 25 times faster than Nvidia’s GPU technology.
In January 2026, OpenAI inked a separate agreement with Cerebras valued at more than $10 billion.
The chip startup secured $1 billion in funding during February 2026, pushing its valuation to approximately $23 billion.

Amazon Web Services has entered into an extended collaboration with semiconductor startup Cerebras Systems, bringing its Wafer-Scale Engine technology into AWS infrastructure. The arrangement focuses on AI inference workloads — the computational phase where artificial intelligence models generate responses to user inputs.

$AMZN's AWS and Cerebras are partnering to bring faster AI inference to Amazon Bedrock in the coming months.

The setup will combine Trainium for prompt processing with Cerebras CS-3 for token generation, with Bedrock becoming the first cloud service to offer Cerebras’… pic.twitter.com/LT6yoDEmCB

— Wall St Engine (@wallstengine) March 13, 2026

As the world’s dominant cloud computing provider, AWS has traditionally relied on internally developed silicon, particularly its Trainium processors created by semiconductor division Annapurna Labs. This new arrangement will see AWS merge Trainium capabilities with Cerebras technology to deliver enhanced inference performance.

Amazon.com, Inc., AMZN

According to Cerebras, its Wafer-Scale Engine excels during the “decode” stage of inference — the critical moment when models produce their actual output — achieving speeds up to 25 times greater than Nvidia’s graphics processing units.

AWS will market this as a high-performance option. “If you want slow inference, there will be cheaper ways to go,” stated Cerebras CEO Andrew Feldman. AWS representatives confirmed they will continue providing budget-friendly inference alternatives powered exclusively by Trainium.

Cerebras Momentum Accelerates

This partnership arrives several months following OpenAI’s January 2026 deal with Cerebras, an arrangement reportedly exceeding $10 billion in value. That contract aims to supply OpenAI’s ChatGPT platform with Cerebras computing power, with plans to establish up to 750 megawatts of processing infrastructure.

During February 2026, Cerebras completed a $1 billion capital raise, elevating total funding to $2.6 billion and establishing a company valuation near $23 billion. Investment participants include Fidelity Management, Benchmark, Tiger Global, and Coatue.

The company had previously submitted IPO documentation in September 2024 but pulled the application approximately one year afterward.

Nvidia Faces Growing Competition

The AWS-Cerebras collaboration represents another competitive pressure point for Nvidia within the inference computing segment. Industry momentum has been transitioning from model training workloads, where Nvidia maintains market leadership, toward inference applications requiring superior processing velocity.

Nvidia continues advancing its position. During December 2025, the company finalized a $20 billion licensing arrangement with chip developer Groq. Additionally, Nvidia has indicated plans to introduce a novel processing platform incorporating Groq’s innovations in the coming period.

Nafea Bshara, AWS executive and Annapurna Labs co-founder, emphasized that the Cerebras collaboration centers on accelerating performance while reducing expenses. “Our job is to push the speed and lower the price,” he explained.

Amazon’s AMZN stock traded down 0.44% at the time of this report.

✨ Limited Time Offer