Skip to main content

Cerebras Wafer Scale (WSE)

Vendor: Cerebras (IPO completed)

Category: ASIC Dedicated Accelerator

Architecture: CS-3

Introduction

The Cerebras Wafer-Scale Engine (WSE-3) is the largest AI processor in existence, featuring 4 trillion transistors and 900,000 AI cores on a single chip. Focused on large model training, it delivers performance comparable to traditional clusters without requiring distributed programming. On May 14, 2026, Cerebras went public on NASDAQ with a first-day valuation of approximately $560 billion.

Specifications

ModelComputeMemoryInterfaceTDPProcess
WSE-34 trillion transistors, 900K cores44GB on-chip SRAMCS-2 Fabric20000+5nm
WSE-22.6 trillion transistors, 850K cores40GB on-chip SRAMCS-2 Fabric20000+7nm

Official Website

Visit Official Website

Driver Downloads

Linux

OS Support

WindowsLinuxmacOSAndroid
✅ (Cerebras Cloud)

Version History

VersionRelease DateDescription
Cerebras Cloud 2.02024WSE-3 launched + API access

Performance Benchmarks

ModelTaskPerformance Metric
WSE-3Llama 3 70B Inference~1800 tok/s (pipeline)
WSE-3Ultra-long context inferenceSupports 128K context
WSE-2GPT-J 6B TrainingCompleted in minutes

Pricing

ModelReference PriceNotes
WSE-3Cloud API pay-per-useCerebras Cloud / CSP
WSE-3Contact vendorBare-metal deployment

Quick Installation

Cerebras Cloud (API)

# Install Cerebras SDK
pip install cerebras-cloud-sdk

# Use Cerebras Inference API
# OpenAI API-compatible interface

WSE series is primarily accessed through Cerebras Cloud API or partner cloud services. Bare chips are not sold to end users.

Code Examples

Python (Cerebras API)

from cerebras.cloud.sdk import Cerebras

# OpenAI-compatible interface
client = Cerebras(api_key="your-key")
response = client.chat.completions.create(
model="llama3.1-8b",
messages=[{"role": "user", "content": "Hello"}],
max_tokens=100
)
print(response.choices[0].message.content)

Architecture Highlights

  • Wafer-Scale Engine (WSE): An entire wafer as a single chip. WSE-3 features 4 trillion transistors and 900K AI cores
  • CS-3 System: 16 WSE-3 chips interconnected, supporting trillion-parameter model training
  • Zero Fragmentation: Wafer-scale design eliminates traditional inter-chip interconnect bottlenecks with ultra-low latency

Model Compatibility

Model/FrameworkSupportNotes
Llama Series✅ NativeOfficial Cerebras deployment
Large Language ModelsAPI inference
Training Frameworks⚠️Via Cerebras PyTorch extensions
Custom Models⚠️Requires Cerebras customization

Large-Scale Cluster Deployments

Based on global AI supercomputing cluster statistics, Cerebras WSE has accumulated over 65 chips deployed across 2 publicly disclosed clusters.

Chip Model Statistics

Chip ModelTotal DeployedCluster Count
Cerebras CS-2641
Cerebras CS-111

Notable Deployment Clusters Top 10

#Cluster NameTotal ChipsChip ModelOperator
1Condor Galaxy 1 (CG-1)64Cerebras CS-2 ×64G42,Cerebras Systems, United States of America
2Lawrence Livermore NL Lassen Phase 21Cerebras CS-1 ×1US Department of Energy, United States of America

If you're evaluating alternatives, the following products may also fit your scenario: