Next Generation IPU Systems
IPU-M2000 + IPU-POD4
Core building blocks for AI infrastructure at scale
Buy Now
The IPU-Machine: IPU-M2000
The IPU-M2000 is the fundamental compute engine for IPU-based machine intelligence, built with the powerful Colossus Mk2 IPU designed from the ground up for AI. It packs 1 petaFLOP of AI compute with 3.6GB In-Processor-Memory™ and up to 256GB Streaming Memory™ in a slim 1U blade. The IPU-M2000 has a flexible, modular design, so you can start with one and scale out to many in our IPU-POD platforms.
Directly connect one IPU-M2000 to a host server from our OEM partners, to build the entry level IPU-POD4 to start your IPU journey.
Read Analyst Report
Graphcore IPU-POD4
Get started with the IPU-POD4 - the entry-level AI compute engine to start your IPU journey, combining a single IPU-M2000, directly connected to a host server, in a 2U pre-qualified system.
Watch videoPerformance
As the fundamental compute engine for all our IPU-POD platforms, the IPU-POD4 - one IPU-M2000 plus your choice of approved host servers - delivers real performance advantage against the A100 GPU in the DGX platform.
Finance firms achieve blisteringly fast time-to-result seeking alpha with Markov Chain Monte Carlo models.
Search engines get results faster for their users using Natural Language Processing models like BERT-Large.
Markov Chain Monte Carlo: Training
.svg "MCMC")
Click to Zoom
BERT-Large: Inference
.svg "Bert Large")
Click to Zoom
Inside the IPU-M2000
.svg "IPU-M2000 Diagram")
READY TO GET STARTED?
Connect with our experts to assess your AI infrastructure requirements and solution fit.
×