The scale of Intel’s AI chip enterprise as we speak is big, but it surely’s nothing in comparison with the place it expects to be in simply 4 years’ time.
Talking throughout the firm’s Innovation Summit in Santa Clara, Intel Govt VP Navin Shenoy revealed a brand new deal with AI growth.
The corporate’s AI-focused Xeon processors generated $1 billion in revenues throughout 2017. By 2022, it expects to be producing round $10 billion per yr.
AI is ready to be carried out in lots of areas of our lives within the coming years, throughout a wide range of units.
Shenoy claims current breakthroughs have elevated the corporate’s AI efficiency by 200x since 2014. He teases additional enhancements are on their approach in upcoming releases.
The corporate will likely be launching its ‘Cascade Lake’ Xeon processor later this yr with 11 instances higher efficiency for AI picture recognition.
Arriving in 2019 will likely be ‘Cooper Lake’ which makes use of 14-nanometer manufacturing and can characteristic even higher efficiency. In 2020, nonetheless, the corporate is focusing on ‘Ice Lake’ with 10-nanometer manufacturing expertise.
“After 50 years, this is the biggest opportunity for the company,” says Shenoy. “We have 20 percent of this market today.”
The admission it presently has a small share of the market as we speak is daring and exhibits the corporate is assured about considerably upping that share within the coming years. It faces vital competitors from Nvidia particularly.
Intel’s revenues had been round a 3rd data-centric 5 years in the past. Now, it’s round half of Intel’s enterprise.
Shenoy’s feedback as we speak present how significantly Intel is taking its AI enterprise and the agency’s confidence it will likely be a serious participant.