I’ve looked into the opportunity NVDA has in Automotive. If you’re investor, you really should spend the time listening to the Company’s presentations given on Investor Day on May 10, 2017. The webcast is archived at the following page on NVDA’s website:
The Investor Day was held during the GPU Technology Conference. There was a presentation given by Rob Csongor their VP of Automotive. Here are some highlights:
Complete computing stack (open source) for automotive
245 automotive companies at GTC; all using the computing solution and AI stack
35 vehicle makers (20 car makers); John Deere, forklifts in warehouses, and other types of vehicles will use the technology
24 tier 1s
10 HD mapping companies (announced a partnership with every single mapping company)
Launched Drive PX2 in last year
Q4 2017 introduced TensorRT AI compiler
Companies developing with AI in Automotive jumped from 60 to 170 companies
End of year will launch Xavier which will integrate into one chip (from currently 4 processors: 2 Parkers, 2 GP106s)
The above tells me that NVDA is trying to make it easier and easier for companies to develop AI for Automotive. It’s a pretty big focus for NVDA. It’s also evident that NVDA has only recently enabled companies to easily use NVDA’s products and solutions for Automotive. Here are the revenues for Automotive again:
During Q1 2018, the number of partners jumped from 60 to 170! What will that mean for revenues going forward? I think it’s likely that we will see revenues to the Automotive segment accelerate. Toyota is probably their largest partner right now but they are just one partner of 170. Players in the car industry usually add innovations that one player has introduced. If Toyota introduces level 2 autonomous driving into all their US and Japan cars then the other car makers will also. Their Xavier chip that will launch later this year will enable more applications and AI development; it should also contribute to further accelerate revenue growth for NVDA in the sector.
If Toyota introduces level 2 autonomous driving into all their US and Japan cars
A big IF for cautious Japanese car companies. Doing this on all their cars would substantially increase car prices, disproportionally for the non luxury models. Which accounts for a lot of their sales. These cars would then sell for more the their rivals. Most of which are near clones.
I doubt if they will take this Musk type gamble especially since their polls show most Toyota drivers are not interested. Nobody buys a Toyota to be on the cutting edge to automotive technology.
I doubt if they will take this Musk type gamble especially since their polls show most Toyota drivers are not interested. Nobody buys a Toyota to be on the cutting edge to automotive technology.
One must be careful when making this assumption. Polls can be misleading. If the poll asks “Would you pay an extra $200 for level 2 automation?” many might answer no. If you ask a different question such as “Would you pay $200 if your car could drive itself in bumper to bumper traffic?” then you might get a difference response. Even so, getting an answer from prospective customers on what they want is often very difference from the actual purchasing behavior.
If you read the Barron’s article, there is explicit confirmation that Needham met with Toyota who confirmed that they will be aggressively deploying level 2 advanced systems into the mass market in the US and Japan over the next 2.5 years. Will Toyota actually do what Needham says Toyota told them? That seems more certain to me than Mauser saying he doubts it. Mauser, can you provide the link of the Toyota polls?
then there is the big future switch to fuel cells by Toyota. Told to analysts as well as public .Never happened OIMO it was a deliberate attempt to shift attention away from BEV.
No link but there have been published reports of polls done by others Where consumers are mostly uninterested in advanced car automation.
http://www.wbur.org/bostonomix/2017/05/25/mit-study-self-dri…
there are lots more- google it. people do want some autonomous technology in their cars — such as automatic emergency braking, lane-keeping systems and auto-park features. “They’re looking for driver assistance systems that work to help them stay in active control [and] safe control of the vehicle,” Reimer said. "They’re just not looking for a car to drive them in a chauffeur kind of framework.”
So maybe L2( but not much beyond that ) is wanted by(some) consumers in which case Toyota would be following the crowd.
Agree you can mostly get whatever answers you want in a poll.Which is why all the autopilot polls don’t give the same answers.
Maybe L2 won’t cost much by then,. It also does not do much.Tesla will be well beyond that by then.
I don’t know what Toyota will do, only that they are not always honest. And they could change their mind. A question from Nvidia point of view is whether they will use Nvidia technology and if so how much. IMO they are unlikely to do anything increasing base price by more than a few hundred dollars at most. It’s just too risky.
Time will tell, but whether Toyota does or does not will not influence my stock selections today. All car companies will have to add it, the question is “When” and" Can Nvidia Hold the Lead."
Thanks for the additional info on Toyota. Definitely need to take what they say with a grain of salt. However, level 2 driving is is very small step for consumers compared to fuel cell cars for which there is no infrastructure.
They’re just not looking for a car to drive them in a chauffeur kind of framework.”
So maybe L2( but not much beyond that ) is wanted by(some) consumers in which case Toyota would be following the crowd.
But it is all relative the concepts people have constructed in their minds. When I saw cars that could parallel park themselves (starting at the high-end), my construct was to think it is a stupid waste of money, why don’t you idiots just learn to park for free! But it is working its way down to cheap old Buicks and it will be “free-ish” soon. I might not be able to buy my next car without it.
Auto-braking, lane assist - those are easy and safe constructs. Who wouldn’t want brakes that might sacve your life. But once you are use to that, and lane assist, and L2, then your brain is ready for the next level and you don’t conciously know it. People will see the evolution on Investing channels, then regular news, then all over. They will see the dude in the next lane eating a hoagie with 2 hands and waiving at them. Our brains will evolve. My wife thought microwaves were stupid when they first came out - she could heat water on a stove. But once you have one heating your coffee was convenient. And heating foods was just fine. Cooking microwave designed food on microwave safe plates became common place. Now we are like “why can’t we have a turbo freezer to make things instantly cold like we make them instantly hot”. Our brains change.
Remember how dangerous elevators were? You had to hire someone to “drive” it all day, now everyone is compfortable pushing their own button. Cars are next. It has all moved faster than I expected it and with Google, Tesla, Uber and others pushing, I suspect it will happen faster than we think.
Sure my grandmother had a rotary for a LONG time, but most people moved to push button and wireless.
True Tinker…as far as aurtonomous is concerned, the easiest type of driving would be the same streets, same route day after day…from a programming and sensing perspective. I am not so sure that more expensive and sophisticated algothyms are required for that type of driving wherein there would primarily be changes in traffic and pedestrian issues to contend with.
That is a far cry from the sophistication of varying routes, conditions, traffic, and changes in these issues in real time…very complicated stuff.
As the technology seems to be today…truthfully,…would you drive a Tesla and take your hands off the wheel and read the paper??
I like NVDA, but let’s be clear that Automotive revenue of $0.6B is a drop in their overall $6.9M revenue bucket. It’s not clear how quickly that will grow.
During Q1 2018, the number of partners jumped from 60 to 170! What will that mean for revenues going forward?
Near term it means nothing. You want to play with NVDA chips? You become a partner, buy a dozen and put some engineers on it.
What’s more important for revenues is who is going to production with NVDA? Audi just recently announced that next year’s A8 will have Level 3 Autonomous Driving capabilities power by Nvidia, and has something like 6 chips total. It should be noted that Audi has been using Nvidia chips in its cars for years now, so one model of car hitting next year won’t move the needle that much by itself.