What’s The Deal With Intel?

Intel paid more for Mobileye than Fiat paid for Chrysler. That is $15
billion vs. $4.35 billion. Chrysler makes cars, trucks, SUVs, JEEPS!
Mobileye at least was/is profitable. It is a lot of money.

The only reason companies make such large purchases is because they are
making bet on their future. Like GM’s purchase of Cruise Automation and
Ford’s investment in Argo AI.

These companies all want to shift from
simply making hardware to providing services.

It is a good strategy if you believe in the future of AV and
Transportation-as-a-Service. I’ve publicly spoken often about this
future and why I think it is reasonable.

There is the obviousness that the current reality of manually driven
vehicles suck no matter which metric you use:

  • Fatalities (over 4 million in the U.S. annually)
  • Traffic (costs the U.S. $130 billion annually)
  • Maintenance (costs the average vehicle owner $10,000 annually)
  • Utilization (less than 5% a day)
  • Environmental (do I really need a quote here? And EVs suck as well since
    batteries are not easily recyclable and most electricity is produced by coal)

Larry Burns, McKinsey, Bloomberg, KPMG, and Barclays have all provided
various estimates that autonomously driven, shared vehicles cost per mile
driven may be as low as $0.08 vs. the current $1.60. Them thar’ economics
alone will drive demand and supply. Or is it supply and demand? Let’s
agree it’s both.

CB Insights just released their latest report on automotive technologies
and they point out that in 2016, there were about 100 deals done worth
over $1.1 billion with 208 unique investors. This year has seen 49 auto
tech deals already worth over $1.3 billion. Clearly investors think the
AV and TaaS future worth putting down money.

Given all these reasons, it seems perfectly reasonable that a hardware
company like Intel will make the move from hardware only to moving up the
value stack by adding services.

Intel has a history of making a lot of investments and I honestly do not
know if they fully utilized. I know that Harman had a history of buying
and selling companies, but did not always make the best use of the
technology or people. Any micro economist will tell you, it is very hard
to fully integrate and reap the complete benefits of any purchase. Most
mergers and acquisitions fail within three years to sustain any boost in
profitability and/or stock price.

What I’m fascinated about Intel, is how many interesting pieces of the AV
and TaaS puzzle they own or are invested:

  • They have CPUs for infotainment, ADAS, cellular, and sensors (Infineon,
    Mindspeed, Altera, itseez, Yogitech)
  • Despite Nvidia’s onslaught of marketing in automotive (thanks to Danny
    Sharpio!), Intel does make GPUs, processors for neural nets, and AI
    training (Nervana, Movidius, Xeon Pi)
  • They have LiDar, cameras, and sensors (besides Mobileye, they own
  • They have security systems in McAfee or at least are somehow still related (?)
  • Mapping technologies (Here and Mobileye)
  • Operating systems and platforms (WindRiver, Perrone Robotics)

They have even created a separate autonomous driving lab in the Silicon
Valley headed by Doug Davis, a really cool fellow who I am personally glad
did not fully retire from the company.

Intel is missing some key ingredients:

  • They do not have communications capabilities, no real connectivity
  • They are missing some key software for device management
  • They don’t seem to be in the pathfinding or control systems solutions
  • And at the end of the day, Intel is not a service company

Intel may not have a culture of providing services to companies and definitely not to
end consumers.

Why does this matter? Intel has always been a hardware company and they
have extracted great economic benefit and fabulous share value. The move
to Transportation-as-a-Service is going to require a blending of hardware,
software, IT platforms, and services. I think the real returns will be on
companies that shift from providing only hardware to providing a service.

Intel’s endless reorganization speaks to a company that is trying on a
bunch of different ideas on for size, but shows a lack of a cohesive
vision and across-the-board buy-in. If BK keeps talking about the
importance of 5G for the future of ADAS, AV, and TaaS, then where are
Intel’s partnerships or purchases of wireless capabilities beyond chips
for radios? Why didn’t they buy Jasper or Cubic or even the company I
work for, Aeris? Without fully understanding and solving wireless in its
entirety, they are leaving money on the table. Connectivity is key to
providing services in TaaS.

It seems like Intel is trying to own a large stack of the solution, but
are missing some very crucial aspects. If they fill in those gaps or get
close, I think it perfectly reasonable they would spin off a TaaS company
to tackle the intricacies they do not fully understand or cannot complete
under their current, albeit ever-shifting, organizational structure.
Maybe they would form a separate LLC like Ford’s Smart Mobility or Toyota

One of my biggest worries is about automotive product planners. When I
worked at Toyota, it always baffled me when product planners would source
the smallest amount of CPU power for their vehicles, never allowing for
future power to be used by future services. There was a sense that less
is more and more is waste. Long gone are the days when I was close to CPU
product roadmaps and daily discussions with engineers as to why they are
choosing each chipset. I genuinely worry that the next five years of
vehicles are being planned with no spare capacity and that in and of
itself will stymie progress.

It would suck to have all our future’s possibilities cut short by myopic
engineers doing business as usual while the world around them, even Intel,
tries to make things better. And by better I mean safer, easier, more
widely available, and less expensive.