Is Tesla's (NASDAQ: TSLA) AI Chip the Ultimate Edge?

Dear Investor,

It would suffice to say that things have been better for Elon Musk.

At the top of the billionaire entrepreneur’s list of woes is the SEC. Tesla just forked over $20 million to the securities regulator as part of a settlement surrounding a misleading Tweet by Musk. Upon hearing the news, we couldn’t help but think, “Only in 2018.”

Then there was his participation in a recent podcast produced by Joe Rogan. You can listen to the whole podcast on YouTube — all two hours and 37 minutes of it. It’s a whirlwind conversation wherein Musk warns about the future of AI, hints at connecting human brains to computers, and, to the dismay of Tesla shareholders, takes a drag from a joint.

The marijuana use was what made headlines, of course, and short sellers laughed with glee when they heard the news. Unfortunately, it also indicates that one thing bulls and bears alike didn’t do was listen to the podcast: Despite sipping whiskey in a laid-back atmosphere created by Rogan, Elon was very much himself.

That’s not to say shareholders shouldn’t be concerned. On top of the aforementioned controversies, Musk’s personal life doesn’t seem to be great, either. In an August interview with the New York Times, Elon got teary-eyed when talking about his inhuman work schedule and public breakup with musician Grimes.

That said, investors shouldn’t get too distracted by the tabloid headlines and Elon’s personal life. At the end of the day, it’s the company that matters, which is why we’re paying attention to something very big that just happened over at Tesla HQ.

Tesla Has its “Apple Moment”

Lots of people like to throw around “Apple moments” when talking about Tesla. We count at least half a dozen.

The one we’re referring to is Apple’s recent decision to drop Intel. The reason is speed. Mobile computing technology continues to grow and evolve — so much so that the old “Intel way” of doing things (updating chips every few years) just won’t cut it anymore.

In order to stay ahead of iPhone competitors, Apple had to start making its own chips.

And so it is with Tesla. Previously, Tesla’s autonomous driving AI chips were powered by NVIDIA (NASDAQ: NVDA) GPUs, but that will soon end. For close to two years now, Tesla has been working on building its own computer chips — a feat for anyone, let alone a car manufacturer.

As first reported by Electrek, it began in 2016 with the hiring of microprocessor engineer (and creator of Apple’s Ax chips) Jim Keller as Tesla’s Vice President of Autopilot Hardware Engineering. The reason for hiring Mr. Keller was in his title: They were trying to build autonomous driving capabilities and, in true Tesla fashion, had decided to build the chip themselves.

Keller was the perfect hire because, in addition to formerly working at Apple, the chip engineer was fresh off a three-year stint at AMD where he led the Zen processor development team.

Of course, Elon Musk didn’t admit WHY he hired Keller. At least not at first. In December 2017, he spilled the beans: Keller was there to develop a chip optimized for self-driving. The venue for such an announcement, the Conference on Neural Information Processing Systems (say that five times fast), was naturally abuzz.

At the time, Musk said that when finished, he expected Keller to bring “the best customer AI hardware in the world” home for team Tesla.

Surprisingly, Keller left Tesla at some point in the past year, but not before getting Tesla well on its way to building a fancy new custom AI chip.

During Tesla’s quarterly earnings conference call in August, the AI chip team’s new head, Pete Bannon, went on record saying that Tesla had “the world’s most advanced computer for autonomous driving” and would be rolling it out to current Tesla owners in 2019.

Maybe Keller is like the hero in old Westerns, riding into town only to ride off into the sunset having done his duty.

Are You John Connor?

And boy, did Keller do his duty.

On October 16, 2018, the tech world was rocked with the news that Tesla will be rolling out the “world’s most advanced autonomous driving neural net computer” in April 2019.

All Tesla owners that ordered full self-driving will be getting the chip upgrade to their vehicles free of charge. How’s that for customer service?

Naturally, Elon Musk wouldn’t be Elon Musk if he didn’t make the revolutionary announcement via Twitter.

No doubt this is a big deal in the emergence of autonomous driving technology. And it certainly lends itself to the bull case that Musk is light-years ahead of the automaker “old guard” in both electric car manufacturing and autonomous driving technology.

The new chip will be able to handle 2,000 frames per second, with redundancy, while it’s processing data from a unified eight-camera network.

For perspective, that’s anywhere from a 500% to 2,000% increase in computing power from the NVIDIA GPU-powered Autopilot 2.0 Tesla hardware.

What Does it All Mean?

All of this is great news for Tesla shareholders and Elon Musk fans. We, however, are more interested in its implications for electric cars and autonomous driving.

Other automakers, we believe, will eventually try to replicate what Musk is doing. The one sure thing is that driverless advances within the larger trend of electric vehicle growth are more a question of when, not if.

This has implications for cities, national governments, investors, and the companies that produce the raw materials that will make all this possible. To find out how to position your portfolio for the inevitable rise of EVs, click here.

Until next time,

The Wealth Daily Team