There’s a bit of trend going on at the moment in terms of some of the larger chip manufacturers turning to the GPU for their new designs.
However, European-based startup Graphcore takes a different route in looking at how the processing for artificial intelligence (AI) should be.
While Graphcore doesn’t have a product to demonstrate just yet, CEO Nigel Toon expects to have one for early-access customers by the early part of next year.
Even that’s enough to get other companies to design their own GPU technologies to try and incorporate this kind of streamlined processing within language recognition, computer vision, and others the focus on using machine learning.
“What this really does is allows us to scale,” said Toon. “We’re already working on a roadmap, we can tack on an drive the development of those really quickly.
We can look at some other areas, we can expand so we can support more customers more quickly. I think it really allows us to fundamentally speed up.”
The product to be released by the company next year is being dubbed the “intelligence processor unit” currently, or IPU for short.
This AI technology has been designed to be able to cope with all the rapid-fire calculations that are required for machine learning.
Once released, it’s likely to be seen in various devices all over the world that need high powered machine learning operations in which to run.
Having just raised $50 million in a new funding round led by Sequoia Capital, it seems Graphcore is raring to go.
But Graphcore may have some stern competition on its hands.
Other companies including Google and Apple are also looking to build technology that is specific to their needs.
Even Tesla is rumored to be producing its own AI chip in conjunction with AMD.
There are definitely some changes on the horizon in terms of companies developing their own AI chip technologies, but who will get in there first still remains to be seen.
“Having [Sequoia Capital] in, it’s really going to allow us to build a big company, which is fundamentally what we’re hoping to do,” said Toon.
“This is a massive opportunity. This is the next generation of computing. This is the opportunity for a new player to build an industry standard. I see a strong parallel with what ARM was able to do in the mobile space, but I think the opportunity here is really bigger.”