Nvidia boss: cloud, ¡Si! Intel, ¡No!
The times they are a-changin' — big time
Intel: shook-up in the shake-out
Another future terra incognita will be the computing industry. As Huang sees it, "There is no question that ... over the course of the next five years we will see an enormous disruption in the computer industry."
Part of that disruption will be in the way we buy — or even if we buy — our future computing devices. A customer will have to ask him or herself, Huang said: "'Do I buy a computer or do I sign up for a service?' Notice that question that I ask, if it lands in one place or in the other place, completely changes the computer industry. Completely."
That transformation has already, well, transformed the personal handset market, and it's about to do the same for the personal computer market, Huang insists: "Everything we know about the computer industry will change as a result of that question. I don't remember in the last thirty years when this much change is about to happen."
The PC market is about to go through massive change. "The PC is evolving," Huang asserted. "The PC used to stand for personal computing, now it stands for PC. Now there's a new word for personal computing." And although Huang didn't supply that exact word, his comments made it clear that he was talking about devices that are mobile, primarily visual, and highly dependent upon parallel computing.
Those devices, coupled with the cloudy compute capabilities Huang envisions, will set the PC market on its ear. "The future of the personal computer industry is going to be radically different than the PC industry today," he said. "There'll be new OEMs. There'll be new distributors — meaning you won't just buy from Best Buy, you'll obviously buy from Best Buy, Verizon, AT&T, and others."
Finally, Huang predicted that in his vision of the radically different future, "There's going to be a new ISA, a new instruction set architecture."
And, no, he wasn't talking about his company's CUDA parallel processing architecture, he was talking about ARM: "ARM — you can't say enough about this company, right?"
After waxing nostalgic about ARM's history, he said: "Up to today, they've shipped some 20 billion cores. But over the course of the next 10 years they're going to ship another 80 more billion cores. And so you can tell that this is a company that is going somewhere."
Which is a heartwarming sentiment for those Cambridge, England, chip designers, but Huang had another point to make: "The important observation that I think is obvious now to everybody [is that] ARM is the most important instruction-set architecture of the future. ARM is the most important CPU architecture of the future. It's the fastest-growing architecture of the future."
What about Intel, though, Huang was asked, how would he describe Nvidia's relationship with Intel? His answer: "Well, we're trying to have a lesser one."
As Huang sees it: "The computer industry has a hard shell around it. And this hard shell is the traditional PC industry. And that hard shell is occupied by some very, very powerful forces" — and everyone in the conference audience knew that Intel was in Huang's group of powerful forces.
"That hard shell is only penetrable with very, very sharp technology," he said. "And that's what GPU computing was intended to do, and I think we're doing it."
In a few years we'll see if Huang is correct. And if he is, The Reg is certain that he'll be more than happy to lend us his supercomputer-on-your-head binoculars to see Intel receding into the distance. ®
When Huang was asked what kind of personal consumer electronics he'd like to own, he said: "I'm dying to have Tegra phones, Tegra tablets, Tegra settop boxes, I want to be a Tegra home. I've invested well over a billion dollars building it, now I'd like to enjoy it."
Sponsored: Benefits from the lessons learned in HPC