Concept

Graphcore is changing the way machines operate by altering the way they learn. The market for AIs is growing, pushed in part by the development and use of self-learning machines. Graphcore aims to make the machine learning process even more efficient and intelligent. Their technology not only reduces the cost of improving AI, but also helps to incorporate the technology in low-power devices, and speed up the process bringing useful deep learning applications to market.

How they’re disrupting

Being completely different to anything that has come before it, machine learning needed a new technology to fully unlock its potential. Graphcore has built a new processing unit specifically for AIs, allowing them to work faster, smarter and with ease. Focused on “graph-centric intelligence tasks”, this processor is completely new, and will speed up AI development in the emerging machine intelligence market.

Disruption potential

As the AI market is still building, there is an opportunity to take a commanding role in providing new solutions to new needs. According to analyst firm Tractica, the machine intelligence chipset market revenue will increase from $513 million in 2016 to $12.2 billion by 2025. In a competitive new market, the ability to improve machine learning at a faster rate will be a valuable one.

Investments and future

Graphcore’s series A funding round came to $32 million in October 2016. Funding was led by Bosch, with other backing from a range of big technology firms and leading venture capital funds, including Samsung, Amadeus Capital Partners and C4 Ventures. Commercial products will be made available to consumers later in 2017.