But can the chip giant become a serious contender in the autonomous vehicle market?
Intel today said it would acquire embedded vision leader Mobileye for roughly $15.3 billion in equity—$14.7 billion in “enterprise value”—setting the stage for a huge push by the chipmaker into the autonomous driving market.
Intel has been dabbling in the automotive market for some time, starting with an unsuccessful bid to replace 8-bit microcontrollers with low-end processors. With the acquisition of Altera, Intel acquired an important piece of the puzzle, namely the ability to future-proof and customize some of these solutions as they continue to evolve. But the addition of Mobileye rounds out the company’s portfolio with a strong relationship with Tier 1 suppliers, boosting Intel’s standing in this market and setting the stage for an all-out war in this market.
“This helps get Intel a seat at the table and allows them to be engaged with autonomous driving,” said Jeremy Carlson, principal analyst at IHS. “Intel signaled its intent to engage in this space when they agreed to work with Mobileye to develop platforms. If you look at its recent acquisitions, many of those apply to how it will engage in this market.”
Kurt Shuler, vice president of marketing at Arteris, said the deal allows Intel to offer a complete system, from Xeon-level processing to customized processing. “That’s super-compelling for automakers,” he said. “And this is not just about automotive. This technology will be in eyeglasses in the future. It will provide facial recognition and tell you the last e-mail you sent to that person. It will recognize your face when you come to the door, and it will let the police know when a person you don’t know comes to your door.”
Shuler also pointed to a hidden gem in this deal—Mobileye’s people and culture of innovation. “Mobileye is expected to keep all of its people, including Mobileye sales and application engineers. They will now be Intel sales and application engineers. That will jumpstart Intel.”
Intel has other pieces to make all of this work, too. In addition to its ASIC and FPGA chips, it has 3D-Xpoint memory, real-time operating systems, and a long history in servers. As a presentation by Intel to investors argues, cars are evolving into data centers on wheels, with L3/L4 cache adoption expected to surpass 30% by 2030. Intel projects the market for autonomous vehicle systems, data and serves could be as large at $70 billion by 2030.
Semiconductor content in autonomous driving is rapidly evolving into a battle of the deep-pocketed giants, where the table stakes are investments worth billions of dollars. Nvidia, Xilinx, Qualcomm (especially with its pending purchase of NXP) and Infineon all are vying for a major piece of this market, which has unfolded far faster than anyone expected. Tesla’s announcement last October that all its vehicles now have full self-driving hardware greatly accelerated this race, ramping up the competition to a new level.
Intel began stepping up its footprint in this market last year, using Mobileye as the common denominator. Intel teamed up with Mobileye and BMW last July. In January, the three companies announced that about 40 autonomous BMWs would be on the road by the second half of this year. Intel also formed an alliance with Mobileye and Delphi in November to develop autonomous vehicles.
By acquiring Mobileye, Intel also gains a development partner that is outside of Intel’s culture of focusing on ever-faster processing platforms to fit all needs. Mobileye, based in Israel, develops integrated hardware and software systems. Whether Mobileye will be kept at arm’s length, similar to carmakers’ strategy to locate offices in Silicon Valley, or whether it will be incorporated into Intel remains to be seen. But the headquarters of the the acquired group, at least initially, will remain in Israel under the leadership of Mobileye Chairman Amnon Shashua.
In an e-mail to Intel employees, CEO Brian Krzanich wrote, “The acquisition of Mobileye brings together the assets of Intel’s Xeon processors, FPGAs, 3D XPoint memory, and 5G modems with the world leader in automotive computer vision. This acquisition essentially merges the intelligent eyes of the autonomous car with the intelligent brain that actually drives the car.” He added that the reason why autonomous vehicles are so important to Intel is “DATA…We are a DATA company.”
But Intel’s costly failure in mobile phones is a stinging reminder that its PC-era culture doesn’t play well in all markets. The question now is how well Intel can adapt to a variety of completely different markets.
Seeing The Future Of Vision
Self-driving cars and other uses call for more sophisticated vision systems.
Teaching Computers To See
First of two parts: Vision processing is a specialized design task, requiring dedicated circuitry to achieve low power and latency.