Friday, January 6, 2017

Breakfast With NVIDIA's CEO: "‘Reflecting on Destiny Meets Serendipity’" (NVDA)

Tiernan Ray at Barron's Tech Trader Daily brings in the story:
Jan. 5, 2016     12:40 P.M. ET
Thursday morning, Nvidia (NVDA) founder and chief executive Jen-Hsun Huang followed his keynote at the Consumer Electronics Show the night before with a breakfast Q&A for a few journalists and I was fortunate enough to be invited.

Huang’s most interesting statement about the big picture came at the end of the talk, when he reflected on Nvidia’s success at the moment in “machine learning” and artificial intelligence, characterizing it as “destiny meets serendipity“:

The destiny part is, we created a processor that is very good at data processing at high throughput. The serendipity part is that machine learning needs a great computational engine, and so when this elegant algorithm came together with the GPU, it was destiny meets some amount of serendipity.

Huang reiterated his three themes of the night before: the video game market is becoming more and more one of the dominant activities of the culture, and the PC is the best way to do it; the home is going to become a center of A.I.; and your car, using lots of Nvidia trips, is going to “become an A.I.,” including autonomous driving, but also the co-pilot functions of monitoring the behavior of the human driver, through things like lip-reading.

Among interesting tidbits Huang offered were that the co-pilot function in the car is harder to do, from a compute standpoint, than is the self-driving function.

“It takes a full Xavier,” he said, referring to the company’s octal-core driving compute chip.
As far as all the other entrants into the self-driving market, Huang boasted of the company’s work for Tesla (TSLA) (Huang himself drives a Model S), and said Nvidia would be ahead of the competition, ultimately, in terms of actual shipping volume of cars with advanced autonomous driving.

He was asked about small companies, and what they can do in this time of cloud computing giants.
“This is a great time for startups,” said Huang. “We are working with 1,500 startups today, we’ve never been able to work with this many in our company’s history.”

Huang predicted the rise of numerous small businesses, really startups, that have particular data, saying they will make a big business out of their data by using A.I. to process and analyze it. “This is a really good time for startups,” he repeated, likening such data-rich companies to micro-breweries.

Asked about how the company will build out the network of GPUs to support its “GeForce Now” gaming service, Huang said there were a number of ways to do it, including co-location of Nvidia’s own gear in data centers....MUCH MORE
The stock was down little yesterday, up a little today (+1.62) at $103.36.