The Graphcore Second Generation IPU

Print Friendly, PDF & Email

Our friends over at Graphcore, the U.K.-based startup that launched the Intelligence Processing Unit (IPU) for AI acceleration in 2018, has released a new whitepaper introducing the IPU-Machine. This second-generation platform has greater processing power, more memory and built-in scalability for handling extremely large parallel processing workloads. This paper, “The Graphcore Second Generation IPU,” explores the new platform and assesses its strengths and weaknesses compared to the growing cadre of potential competitors.

The well-funded startup has a blue-ribbon pedigree of engineers, advisers and investors, and enjoys a valuation approaching $2 billion. Its first-generation hardware is now available on the Microsoft Azure cloud as well as in Dell-EMC servers. Both companies are investors. Graphcore is now betting its future on this second-generation platform, a plug-and-play building block for massive scalability that is currently unique in the industry.

Supported by the company’s Poplar development stack, the new platform highlights the 7nm Colossus MK2 in a four-way IPU-Machine appliance, and can scale up to 64,000 IPUs across 1024 racks. The fully configured AI supercomputer can deliver some 16 exaflops of AI (16-bit FP) performance. As with the first generation, the company’s focus is to simplify high-performance parallel computing at significant scale.

Download this white paper, “The Graphcore Second Generation IPU” to learn about how the new Colossus MK2 IPU in a plug-and-play hardware and software platform provided by Poplar and the IPU-Machine seems like an excellent strategy for wider adoption of Graphcore technology.

REGISTER NOW! Graphcore’s upcoming webinar “Transform Your Data Center with Our IPU-POD Scale-out Technology” on January 27, 2021.