CS

Cerebras CS-4

Hardware

by Cerebras Systems

Rack-scale AI system built on three WSE-3 Turbo wafers per system and the Nexus rack-scale platform, aimed at ultrafast inference for frontier models; Cerebras says first shipments begin this quarter.

Find alternatives to Cerebras CS-4

What it does

• Server systems
• Accelerator silicon

Sources

Pricing

Not found · 18 Oct 2026

Page read, including its datasheet link; it prints no price.

Description

Official documentation · 18 Oct 2026

"Three WSE-3 Turbos per System"; "CS-4 is the first system built on the Cerebras Nexus Platform Architecture."; "First CS-4 shipments begin this quarter."

What it does

Official documentation · 18 Oct 2026

"CS-4 delivers up to 30x faster inference compared to GPU systems"; "Enables rapid deployment in hyperscale datacenters".

Sold within

Not found · 18 Oct 2026

Page read; it states no distributor, reseller or order route, only "Get started" and a datasheet.

Name

Official documentation · 18 Oct 2026

"Introducing the all new Cerebras CS-4, a revolutionary rack-scale solution that delivers up to 30x faster inference compared to GPUs".

URL

Official documentation · 18 Oct 2026

The product's own page, "Introducing the all new Cerebras CS-4"; the Cerebras Systems home page links it with a Learn more button.

Deployment

Official documentation · 18 Oct 2026

"Bring the fastest AI to your data center." The system is installed in the buyer's data centre.

Also from Cerebras Systems

• Cerebras Inference Model API

  Hosted inference service that serves open-weight models on Cerebras wafer-scale hardware, billed per token.

• Cerebras WSE-3 Turbo (WSE-3T) Hardware

  Wafer-scale processor — now WSE-3 Turbo (WSE-3T), four trillion transistors and 900,000 AI cores — sold inside Cerebras' rack-scale systems: three WSE-3T processors power the CS-4, the successor to the WSE-3-based CS-3.

• Cerebras Training Cloud Platform

  Managed training/fine-tuning cloud for models from 1B to tens-of-trillions of parameters without manual sharding or model-parallelism configuration, billed pay-per-hour or pay-per-model (Cerebras-run custom training).

• Cerebras CS-3 Hardware

  Wafer-scale AI system built around the WSE-3 chip for organisations that run their own AI and HPC supercomputer; scales up to 24-trillion-parameter models on a single logical device and clusters from 1 to 2048 systems.

Open Cerebras Systems in the directory

Something wrong here? Send a correction — quote this product id: cerebras-cs-4.