Data Science Investment Counter —Funding raised by UK data science companies in 2018.
£ 5.640 Million

AI Chipmaker Graphcore Raises £38M from Sequoia


A graph showing a machine learning model being trained on a Graphcore IPU

Graphcore, a Bristol-based company whose processors could dramatically speed up the training of machine learning systems, is bound to raise £38.2 million ($50 million) in a Series C funding round by Sequoia Capital.  The company is expected to start shipping  its chips, called Intelligence Processing Units (IPU) by early 2018.

In October, Graphcore shared new data showing how IPU could turbocharge machine learning training and inference workloads by 10 to 100 times, compared with state-of-the-art hardware (that is: Graphic Processing Units or GPU.)

If that is confirmed, IPUs would allow machine learning models to be developed in a fraction of the time it currently takes to train them, and at a much lower energy cost. Graphcore also says the breakthrough would also result in new kinds of artificial intelligence techniques being devised— which could lay down the foundation for “general machine intelligence.”

This new round will help Graphcore —which was founded in 2016 by Nigel Toon and Simon Knowles [read Shack15 interview with Toon here]—meet the global demand for IPUs, by scaling up production and strengthening the community of developers using Poplar, the software platform Graphcore created in conjunction with their processors.

Corporate Photographer London

Graphcore cofounders Nigel Toon (left) and Simon Knowles.

“Efficient AI processing power is rapidly becoming the most sought-after resource in the technological world. We believe our IPU technology will become the worldwide standard for machine intelligence compute.” Graphcore CEO Nigel Toon said in a statement. “ The performance of Graphcore’s processor, compared to other accelerators, is going to be transformative, whether you are a medical researcher, roboticist, online marketplace, social network or building autonomous vehicles.”

“Sequoia has a deep understanding of what it takes to scale and build a successful business and we are looking forward to a long partnership with them,” he added.

Sequoia partner Matt Miller , who will now join Graphcore’s board, also commented on the round: “Machine intelligence will cause an explosion of new applications and services that will transform every industry. The unique nature of this workload and scale of this opportunity creates an opening where a valuable new chip company can be created. We believe Graphcore’s product architecture, team and early market interest make it the best positioned new entrant in this market.”

Graphcore has already attracted interest from a number of AI hotshots, including  DeepMind cofounder Demis Hassabis,  and t Uber Chief Scientist Zoubin Ghahramani, who both invested in the company earlier this year.

Images via Graphcore


Co-working space and blog dedicated to all things data science.

Subscribe to our newsletter