September at Cerebras: Scaling fast with new funding, data center & more

View in browser >

September at Cerebras:

Scaling fast with new funding, data center & more

📈 Series G Funding: $1.1B raised at a record $8.1B valuation

🏗️ Meet OKC: New Oklahoma City data center with 44+ Exaflops

🎟️ Cerebras @ AI Infra Summit: Keynote, announcements & more

🚀 K2-Think on Cerebras: MBZUAI & G42's new reasoning model

👀 And More... podcasts, articles, events and social!

Series G Funding

Today, we announced an oversubscribed $1.1B Series G funding round led by Fidelity Management & Research Company and Atreides Management. The round included significant participation from Tiger Global, Valor Equity Partners, and 1789 Capital, as well as existing investors Altimeter, Alpha Wave Global, and Benchmark — valuing Cerebras at $8.1B.

Cerebras' 20x AI performance advantage over NVIDIA GPUs has fueled record demand for real-time use cases like code generation, reasoning, and agentic AI. This investment enables us to scale even faster through:

Breakthrough AI processors, systems, and supercomputers
Expansion of U.S. manufacturing and global data center footprint
Growth of Cerebras Inference and Training to meet explosive demand

Connect to learn how Cerebras can 10x your AI outcomes in 2025 and beyond.

Learn More

Meet OKC

On September 22nd, we held a ribbon-cutting in Oklahoma City at the grand opening of our newest data center, featuring:

🔹 44+ Exaflops of AI compute
🔹 20-40x faster on AI-native workloads than GPUs
🔹 Direct-to-chip water cooling for unmatched efficiency

🔹 Tornado and seismic shielded with triple-redundant power

This moment in Oklahoma City builds on nearly a decade of relentless innovation. We’ve transformed AI infrastructure across North America—from early deployments in California to high-capacity builds in Dallas and Minneapolis, and additional sites across the continent—raising the bar for performance, reliability, and sustainability.

Read Our Blog

Cerebras @ AI Infra Summit

At AI Infra Summit, Cerebras had a major presence—anchored by CTO Sean Lie’s keynote on why wafer-scale inference is the path to real-time AI—plus a packed showfloor with demos, talks, and interactions.

We unveiled our Cerebras API Certification Partner Program, bringing Cerebras' world record speed to Dataiku, Vercel, Portkey, True Foundry, and existing partners Hugging Face and OpenRouter.

We also announced streamlined public-sector procurement via Carahsoft and a DoD Tradewinds listing, expanding access to Cerebras industry-leading AI infrastructure for public sector and defense organizations.

Connect With Us

589c175e0ab9bde579966fa0978fa3b42e09a5a6-1649x928

K2-Think on Cerebras

MBZUAI and G42 unveiled K2-Think—the leading 32B-parameter open-source reasoning model—running on Cerebras Inference at a world record 2,000 tokens per second. K2-Think outperforms far larger models while cutting latency and cost. It’s designed for chain-of-thought reasoning, agentic planning, and math/code problem-solving, and is accessible via API with a simple endpoint swap.

Learn More

And More...

LISTEN:

Our Best Podcast Yet – Andrew Feldman, Co-Founder and CEO joined Andy Price on the Artisanal Podcast. This episode was a masterclass in relentless innovation and the mindset of a “professional David going after Goliath.” Learn about the passion, focus, and relentless grind that enabled us to build the world’s only wafer-scale chip — and the world’s fastest AI inference.

Watch now!

LEARN:

What is GPT-OSS? Everything you need to know in 2 minutes

MoE Guide Everything you need to scale Mixture of Experts

JOIN US: