New Frontier for AI Chip Startups

New Frontier for AI Chip Startups
Share this:

Intel is a multinational technology corporation that’s had the monopoly on computer chips for quite some time now.  However, with the advancement of artificial intelligence (AI) that could soon be about to change.

As it turns out, AI works really well with computer chips, and for that reason there are currently more than 45 startups working on this kind of technology today.  These chips have the ability to power things such as self-driving cars and already at least 5 of them have raised over $100 million in investments.  Just last year, more than $1.5 billion was invested into chip start-ups by venture capitalists.

“Machine learning and AI has reopened questions around how to build computers,” said Bill Coughran, partner at Silicon Valley venture capital firm, Sequoia.

In an attempt to stay at the front of the race, Intel first acquired AI chipmaker startup, Nervana for a cool $400 million.  Then a second AI startup called Cerebras went and designed an AI chip with the help of five Nervana engineers.  Earlier this year Cerberus hit the $100 million mark in regards to the total amount of funding its raised.

So far Cerebras has been quite cagey about its future plans.  But, what we do know is that its aim is to accelerate the development of new kinds of AI, including advanced conversational bots and automated video and virtual reality (VR) generators.

Building AI chips is no easy or quick task.  Researchers have to sit for hours testing algorithms across these chips.  The designers of the chips are hoping to use AI to streamline this process to take minutes rather than days.  But like Mr Coughran of Sequoia said, “This is not just about building chips but looking at how these chips are connected together and how they talk to the rest of the system.”

But, that’s just one small part of the challenge.  After the neural networks have been trained for a task, additional resources are then need to carry out that task.  Various chip makers are facing the same problem while at the same time pushing to get AI chips into consumer electronics such as cars and mobile phones.

While no one is exactly sure as to how these new AI chips will work, they’re sure to make a huge difference when they are perfected.  But for that I’m afraid, you’ll be waiting until at least next year.

Source NYTimes

Share this:

Leave a Reply

avatar
  Subscribe  
Notify of