Intel decided this would be a good week to shake up the autonomous vehicle industry, so the chip giant plunked down $15.3 billion to buy Mobileye, the Israel-based company that has been developing a range of technologies for the self-driving industry for nearly 20 years.
The deal between Intel Corp. and Mobileye -- the two companies already have a partnership dating back to last year before this week's acquisition announcement -- is a risky one, especially for Intel. At $63.54 per share, Intel is paying a premium for Mobileye, although CEO Brian Krzanich was at pains to paint the deal as part of a long-term strategy. (See Intel Buying Mobileye for $15 Billion.)
The two companies do expected the acquisition to close by year's end once regulators approve.
Beyond the financials, Intel Corp. (Nasdaq: INTC) is making a huge bet not only the autonomous vehicle market, but on the whole notion of its ability to offer cloud computing capabilities to its customers. The cloud, as it turns out, is the engine under the hood of this market.
The data is destined for the cloud and the heavy lifting of the acquisition will be to move it there.
In a talk with financial analysts following the deal's announcement on March 13, Krzanich explained that Intel's data center capabilities were a major factor in the deal. Specifically, Krzanich spoke of collecting data from the vehicles, analyzing it within the cloud and then redirecting it back to the car or truck in the form of more precise maps to ensure safety and reliability.
Figure 1: Intel's Brian Krzanich has placed his chips on cloud (Source: Intel)
Intel already has a 15% stake in location data company HERE, and has already started work on the mapping requirements for autonomous vehicles. However, by combining those capabilities with Mobileye's technology, Intel is looking to push the limits of what data means in the autonomous vehicle market. (See Intel to Buy a 15% Stake in HERE.)
In his comments, Krzanich laid out three uses for the cloud data center for the technology the two companies plan to develop:
The first is way to capture the visual data that the vehicle records, and then offer new services based on that data.
The second is to use the data collected and stored in the cloud to offer a complete autonomous vehicle offering, which is what the two companies are already trying with BMW, Delphi and others.
The third is to re-imagine the data center, including the way it's built, how data flows from the vehicle and then back to it, and offer a much deeper analysis of that data.
"Put just one million autonomous vehicles on the road and you have the data equivalent of half the world's population," Krzanich wrote in a letter to employees on Monday. "This massive amount of data requires all of Intel's assets to provide the cost-effective high performance solutions our customers need."
Charles King, an analyst with Pund-IT research, wrote in an email that Mobileye provides Intel access to several Tier 1 automotive suppliers, and its combination of hardware and software made it an attractive offer. At the same time, Intel's processors and other technologies open up new ways to bring the technology to the vehicles.
"But the evolution of autonomous driving requires a range of robust IT infrastructure technologies, from muscular onboard components to robust wireless networks to high performance computing and cloud capabilities," King wrote to Enterprise Cloud News. "That is precisely what Intel's Xeon processors, FPGAs, 3D XPoint memory, and 5G modem solutions are designed the deliver."
When Krzanich announced Intel's new direction a year ago, following a large number of layoffs, he spoke about the company's shift to cloud, Internet of Things, artificial intelligence, and machine learning. The company referred to it as a "virtuous cycle" and Krzanich used the same language during Monday's call. (See Intel to Lay Off 12%, Focus on IoT & Data Centers.)
The Mobileye acquisition now puts all the pieces in place. The question is can Intel use its know-how, especially in the cloud, to pull all that data together? The next nine months will offer some guidance.
— Scott Ferguson, Editor, Enterprise Cloud News. Follow him on Twitter @sferguson_LR.