When Intel’s chief executive addressed shareholders at last month’s investor day, he insisted that the “traditional market view” of the company as a Silicon Valley stalwart making chips for PCs and servers was wrong.
“The way we look at the world is a bit different,” Brian Krzanich said. “We are a data company.”
Investors would soon discover just how differently Mr Krzanich was looking at Intel’s world. On Monday, he announced the $15bn acquisition of Mobileye, a leader in the fast-growing market for assisted-driving systems — the second-largest deal in Intel’s 49-year history.
“Together, we expect to be the global leader in autonomous driving,” Mr Krzanich said, “and we see this market as growing to an over $70bn TAM [total addressable market] by 2030.”
Semiconductor makers are swarming over the automotive industry in the hopes that the increasing use of electronics in cars will offset a slowdown in smartphones and the PC’s decline.
Last year Qualcomm paid $47bn for NXP and Samsung bought Harman for $8bn, while Softbank’s $32bn acquisition of UK-based chip designer Arm Holdings was also partly motivated by the anticipation of smarter, more connected cars.
The market will be huge and they are in a dominant position
Mobileye, founded in Israel in 1999, has pioneered computer-vision systems that can “see” cars, pedestrians and other objects on the road, then help vehicles avoid them. Its EyeQ chips and cameras are installed in 15m vehicles, in partnership with 25 automakers, giving it a market share of roughly 80 per cent in “advanced driver assistance systems”.
“The market will be huge and they are in a dominant position,” says Mike Ramsey, auto industry analyst at Gartner, the research group.
Together, Intel and Mobileye hope to offer car manufacturers a one-stop shop for autonomous-driving technology.
“Intel has mapping and infrastructure assets in data centres, artificial intelligence and machine learning, which complements our assets in hardware and simulators,” Amnon Shashua, Mobileye’s co-founder and chief technology officer, said on Monday. “If you put all of that together you get an end-to-end solution for autonomous driving — from car to data centre. It is a very powerful value proposition.”
Like Mr Krzanich, Mr Shashua describes his company not as a chipmaker but as a data company. Starting next year, Mobileye plans to use a new generation of EyeQ to build a constantly updated map of the real world from sensors carried by 2m vehicles. This “crowdsourcing”, which Mobileye calls “road experience management”, is one way in which it has evolved to become a “completely different company” in the two and a half years since its $7.6bn initial public offering, Mr Shashua said in January.
Yet despite the hype in Silicon Valley around self-driving cars and the proliferation of companies experimenting with autonomous vehicles, the market remains nascent.
While Mr Krzanich estimates the first driverless cars could be on the roads by the early 2020s, some analysts predict legal and regulatory roadblocks could push that back by as much as a a decade. Mobileye’s revenues are expected to grow from $358m in 2016 to about $500m this year, representing less than 1 per cent of Intel’s $59bn sales last year.
Ambrish Srivastava, analyst at BMO Capital Markets, notes that Mobileye’s enterprise value of 29 times sales compares with the semiconductor industry’s average of about 5 times. “They are paying, in our view, a very high premium for an opportunity that is still developing,” Mr Srivastava says.
Taking into account Monday’s 2 per cent fall after the deal was announced, Intel’s shares are up by 11 per cent over the past 12 months. Over the same period, rival chipmakers AMD and Nvidia have raced ahead, up by more than 400 per cent and 200 per cent, respectively.
Analysts say the steep price Intel is paying — a 34 per cent premium to Mobileye’s closing price on Friday — reflects the chipmaker’s fear of missing another big opportunity, after sitting out the last decade’s smartphone revolution. “This is a bet that they felt they had to make,” says Mr Ramsey.
Other autonomous-driving entrepreneurs say Jerusalem-based Mobileye is ahead of its Silicon Valley rivals because of its ownership of both hardware and software.
They are paying, in our view, a very high premium for an opportunity that is still developing
“They provide both the chips and the application layer on top,” says Rani Wellingstein, chief executive of Oryx Vision, an Israeli company that is developing laser-scanning technology for self-driving cars.
Thanks to Mobileye’s track record with automakers and relationships with top-tier suppliers such as Delphi, Intel is instantly gaining “a front-seat position in the next-generation mobility business”, Mr Wellingstein says.
Until now, many analysts had attributed pole position in autonomous driving to Nvidia, the maker of graphical processing chips beloved by videogamers and programmers of artificial intelligence systems alike. Last year, in a bitter falling-out with Mobileye, Tesla switched to a computer-vision system developed in-house and powered by Nvidia’s processors for its Autopilot assisted-driving system. In January, Nvidia struck new partnerships with Audi and Mercedes-Benz.
Mobileye faces further competition from top-level auto suppliers such as Bosch and Continental as well as Alphabet’s self-driving car unit Waymo.
Mobileye’s Mr Shashua, however, has been dismissive of new competitors from Silicon Valley. “I meet these people only at CES,” he said at this year’s Consumer Electronics Show in Las Vegas. “I don’t meet them in the marketplace.”