Amazon to Launch AI Models Powered by Cerebras’ Large-Scale Chips on AWS

• March 14, 2026

Amazon Web Services (AWS) is preparing to introduce a new service for running artificial intelligence models utilizing advanced hardware from Cerebras Systems Inc., a startup known for its large-scale AI chips. This move aims to enhance AWS’s current offerings by combining Cerebras’ technology with Amazon’s own Trainium processors to better support large language models.

Boosting AI Capabilities with Powerful Hardware

The collaboration between Amazon and Cerebras is set to provide AWS customers with access to a computing platform optimized for demanding AI workloads. By integrating Cerebras’ unique chips—which are recognized for their enormous size and high processing power—into its infrastructure, AWS aims to deliver improved performance and efficiency for training and deploying large-scale machine learning models.

AWS currently operates as one of the globe’s leading providers of cloud computing resources, offering numerous tools for AI development and deployment. The inclusion of Cerebras hardware is expected to complement Amazon’s in-house Trainium processors, which are also designed specifically for artificial intelligence applications. Together, these technologies will offer more flexible and powerful options for customers working with complex AI models.

The new service is scheduled to launch in the second half of 2026. While the exact financial terms of the partnership between Amazon and Cerebras have not been disclosed, the alliance underscores the importance of specialized AI hardware in advancing cloud-based artificial intelligence technology.

As AI models continue to scale in size and complexity, the demand for dedicated processing solutions is growing. AWS’s strategy to incorporate Cerebras’ chips alongside its own Trainium processors could set a new standard for AI hardware offerings in the cloud, providing developers and businesses with the tools needed to tackle increasingly sophisticated AI workloads.

This partnership highlights a broader trend in the tech industry where cloud providers are seeking to differentiate their services by integrating innovative hardware solutions tailored for artificial intelligence. With AWS’s extensive customer base and Cerebras’s expertise in chip design, the upcoming service is poised to enhance the efficiency and capability of AI model training and deployment in the cloud environment.

Amazon Web Services will offer AI model hosting using Cerebras’ massive chips alongside its own Trainium processors starting in late 2026.