Based out of Bristol, UK, Graphcore is a semiconductor company that develops accelerators for artificial intelligence and machine learning. The company has announced that it has raised $222M (approx €181M) in its Series E round of funding. With this round, Graphcore is now valued at $2.77B (approx €2.25B).
The raised capital will help the company to support its continued global expansion and further accelerate future IPU (Intelligence Processing Units) silicon, systems and software development.
Investors in this round
The round was led by Ontario Teachers’ Pensions Plan Board. It also saw participation from Fidelity International and Schroders as new investors, as well as existing investors, including Baillie Gifford and Draper Esprit.
This investment brings the total funds raised by Graphcore to more than $710M (approx €578.7M), with the company expecting to have over $440M (approx €358.7M) of cash on hand post-closing to support future growth.
According to Graphcore, the new investors have deep experience in disruptive technology and can support scale-up businesses and underscore the size of the market opportunity that the startup is addressing in datacentre AI compute, as well as its product offerings.
Everything about Graphcore
In a press release, Graphcore says that it is developing – Intelligence Processing Units (IPUs) – a type of specialised hardware designed to speed up AI applications, particularly neural networks, deep learning, and machine learning. They’re multicore in design and focus on low-precision arithmetic or in-memory computing, both of which can boost the performance of large AI algorithms and lead to impressive performance results in natural language processing, computer vision, and other domains.
“2020 has been a big year for Graphcore, with the launch of our Mk2 IPU processor, the GC200 and our datacentre compute systems, the IPU-M2000 and IPU-POD64 for scale-out. Our first published benchmarks for the new systems show them clearly dominating the latest GPU-based setups,” the company mentions on its website.
Graphcore was founded in 2016 by Simon Knowles and Nigel Toon, and its technology is currently being used by companies including J.P. Morgan Chase, Oxford University, and the US Laurence Berkeley National Laboratory.
Earlier in March 2020, the company had raised $150M (approx €135M) in funding to secure a position as the leading pure-play machine intelligence AI processor company. The funding has been used in supporting leading financial asset managers, growth equity funds, and investment trusts.
Earlier this month, the company shared its performance results for MK2 IPU-based machine intelligence systems. The IPU-M2000 system outperformed the Nvidia A100 DGX across the board. The company said, Graphcore customers are already making big leaps forward with the second generation IPU systems – whether they prioritise faster time to result, model accuracy, better efficiency, lower TCO (Total Cost of Ownership) or the chance to make new breakthroughs in AI with the IPU.
In the same month, the company also released Poplar SDK 1.4 and with it the first production version of PyTorch for the IPU. “Known as PopTorch, our connecting library brings together the performance of the IPU-M2000 platform and the developer-ready accessibility of PyTorch,” says the company in its blog post. This tightly coupled solution allows users to run standard PyTorch programs on the IPU by simply changing a couple of lines of code.
Prior to that, three months ago, the company had launched its ‘Graphcore Elite Partner Program’ – a global network of channel partners, serving its customers around the world. The Program includes many of the world’s best-known channel partners, technology distributors and resellers, who supported in bringing to market Grapcore’s IPU-Machine M2000 systems. The IPU-Machine M2000 is a plug-and-play Machine Intelligence compute blade that was designed for easy deployment and supports systems that can grow to massive scale.
In early 2020, the company also partnered up with Dell and announced the availability of the DSS8440 IPU Server, and launched Cirrascale IPU-Bare Metal Cloud, an IPU-based managed service offering from cloud provider Cirrascale.