Tesla promises 'one million robo-taxis' in 2020
Submitted via IRC for ErnestGoesToSpace
Tesla Promises Investors 'One Million Robo-Taxis' by 2020
To kick things off, the company shared that it had built its very own computer for self-driving cars. The neural network chip was built from the ground up; the project started back in 2016. Each computer (which is stored behind the glove box) has redundancy so that if one chip fails, the second chip can take over.
This is the company's first time building its own silicon. CEO Elon Musk was quick to boast that Tesla " which has never designed a chip, designed the best chip in the world."
Musk reiterated what he's said before about the hardware available in Teslas. "All Tesla cars right now have everything necessary for self-driving available today. All you need to do is improve the software."
That hardware includes the company's reliance on cameras and radar. When the subject of LiDAR (Light Detection and Ranging) came up, Musk said "LiDAR is a fool's errand. Anyone that's relying on LiDAR is doomed." He later added that "it's fricking stupid. It's expensive and unnecessary."
Source: https://www.engadget.com/2019/04/22/tesla-elon-musk-self-driving-robo-taxi/
Tesla Vaunts Creation of 'The Best Chip in the World' for Self-Driving
At its "Autonomy Day" today, Tesla detailed the new custom chip that will be running the self-driving software in its vehicles. Elon Musk rather peremptorily called it "the best chip in the world... objectively." That might be a stretch, but it certainly should get the job done.
Called for now the "full self-driving computer," or FSD Computer, it is a high-performance, special-purpose chip built (by Samsung, in Texas) solely with autonomy and safety in mind. Whether and how it actually outperforms its competitors is not a simple question and we will have to wait for more data and closer analysis to say more.
Robotaxis also at TechCrunch.
Related Stories
Lots of companies are working to develop self-driving cars. And almost all of them use lidar, a type of sensor that uses lasers to build a three-dimensional map of the world around the car. But Tesla CEO Elon Musk argues that these companies are making a big mistake. "They're all going to dump lidar," Elon Musk said at an April event showcasing Tesla's self-driving technology. "Anyone relying on lidar is doomed."
"Lidar is really a shortcut," added Tesla AI guru Andrej Karpathy. "It sidesteps the fundamental problems of visual recognition that is necessary for autonomy. It gives a false sense of progress, and is ultimately a crutch."
In recent weeks I asked a number of experts about these claims. And I encountered a lot of skepticism. "In a sense all of these sensors are crutches," argued Greg McGuire, a researcher at MCity, the University of Michigan's testing ground for autonomous vehicles. "That's what we build, as engineers, as a society—we build crutches."
Self-driving cars are going to need to be extremely safe and reliable to be accepted by society, McGuire said. And a key principle for high reliability is redundancy. Any single sensor will fail eventually. Using several different types of sensors makes it less likely that a single sensor's failure will lead to disaster.
"Once you get out into the real world, and get beyond ideal conditions, there's so much variability," argues industry analyst (and former automotive engineer) Sam Abuelsamid. "It's theoretically possible that you can do it with cameras alone, but to really have the confidence that the system is seeing what it thinks it's seeing, it's better to have other orthogonal sensing modes"—sensing modes like lidar.
Previously: Robo-Taxis and 'the Best Chip in the World'
Related: Affordable LIDAR Chips for Self-Driving Vehicles
Why Experts Believe Cheaper, Better Lidar is Right Around the Corner
Stanford Researchers Develop Non-Line-of-Sight LIDAR Imaging Procedure
Self Driving Cars May Get a New (non LiDAR) Way to See
Nikon Will Help Build Velodyne's Lidar Sensors for Future Self-Driving Cars
(Score: 0) by Anonymous Coward on Tuesday April 23 2019, @05:34AM (1 child)
here [google.com]
(Score: 1, Informative) by Anonymous Coward on Tuesday April 23 2019, @07:14AM
https://www.marketwatch.com/story/elon-musk-is-just-another-car-salesman-2019-04-22 [marketwatch.com]
(Score: 2, Funny) by Anonymous Coward on Tuesday April 23 2019, @05:51AM (2 children)
Now I understand why he wants to go to Mars, with all those Tesla investors and FSD buyers chasing him down.
(Score: 2) by c0lo on Tuesday April 23 2019, @06:13AM (1 child)
Not going to work if he swindles the Chinese [wikipedia.org]
https://www.youtube.com/watch?v=aoFiw2jMy-0 https://soylentnews.org/~MichaelDavidCrawford
(Score: 2) by GreatAuntAnesthesia on Tuesday April 23 2019, @02:18PM
It's the Libyans [youtube.com] you gotta worry about.
However if he pulls that off then he might just make all his failed 2016 and 2017 deadlines after all.
(Score: 0) by Anonymous Coward on Tuesday April 23 2019, @06:17AM (1 child)
Whose safety? I.e. who will the car kill? The customer and the driver or a dozen pedestrians?
(Score: 0) by Anonymous Coward on Tuesday April 23 2019, @07:14AM
It's a luxury vehicle company for the rich. You do the math.
(Score: 0) by Anonymous Coward on Tuesday April 23 2019, @07:55AM
What's Musk's endgame? Everybody knows his FSD will never work. Stay afloat for as long as possible and hope for a miracle / government handout?
(Score: 2) by RamiK on Tuesday April 23 2019, @08:03AM (4 children)
Last I checked, being the only custom-made autonomous driving chip in the world means it's the best one. No?
compiling...
(Score: 2) by takyon on Tuesday April 23 2019, @09:26AM (1 child)
Nvidia has chips that are custom-made for autonomous driving, last I checked. Drive PX, wasn't it? Oh, it's "Drive AGX Pegasus" now:
https://www.marketwatch.com/story/nvidia-says-tesla-inaccurate-in-self-driving-comparison-2019-04-22 [marketwatch.com]
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 2) by RamiK on Tuesday April 23 2019, @11:22AM
But are they headed to the road by 2020? I'm sure plenty of other companies have development platforms internally, in SBC form or for corporate partners that compete against Tesla's computer. But if Musks-a-lot delivers an on-road autonomous product first, that makes his the first and best driving computer... No?
More importantly, don't forget this is unexplored ASIC: When nVidia is talking about 320 TOPS they're talking about standard neural-nets that they assumed were required separately from Tesla's own development efforts. For all we know Musky is using a custom pathfinding circuitry that nVidia's hardware just does't come near to regardless of the model. I'm not saying it's true. But we just had nVidia's own hardware pulling off raytracing using fairly specialized designs so it's not impossible and we should all start getting used to FLOPS and transistor counts not being the end-all of these sort of discussions.
compiling...
(Score: 0) by Anonymous Coward on Wednesday April 24 2019, @12:46PM (1 child)
Guy only said best chip, not custom-made autonomous driving. So no.
(Score: 0) by Anonymous Coward on Wednesday April 24 2019, @01:48PM
There's no agreed upon qualifiers for "best chip in the world". The fastest ones are made in the hundreds and are kept in labs. The fastest-per-performance are 8$ DSPs that you can't run a toaster on. The best to run windows/linux on are some of the buggiest, most power inefficient pieces of 70s engineering you could come up with...
Overall, if you deny the context of the deceleration, than Musk can just say since it's the best chip Tesla made performance-per-dollar. for a Tesla car and that would be true. A useless metric. But true.
(Score: 3, Funny) by jjr on Tuesday April 23 2019, @12:37PM
So the Johnny Cab was a Tesla after all!
(Score: 1) by Rupert Pupnick on Tuesday April 23 2019, @01:28PM (1 child)
Anyone who understands IC design and manufacturing knows that you’ll never get the same performance per unit area out of an ASIC as a full-custom IC, and both of those technologies fall under the category of “chips”.
(Score: 0) by Anonymous Coward on Tuesday April 23 2019, @06:19PM
> you’ll never get the same performance per unit area out of an ASIC as a full-custom IC
Not sure that the world is quite to that stage of development?
My guess--if Tesla has taken core (and/or bottleneck) portions of their self-driving software and implemented them in an ASIC with minimal overhead, they are going to run a whole bunch faster than on a general purpose processor, and probably faster than on any re-purposed AI engine or GPU.
The problem that I've heard about with the Tesla scheme is that the cameras just aren't very high resolution. The semi that crossed the Florida highway, T-boned by a Tesla, was barely a few pixels in the camera view, until the car was pretty close. Have forgotten details, someone posted images from different distances and human eyes are much better at that distance--an alert driver would have slowed down in advance, anticipating when the semi would finish crossing the highway.
(Score: 3, Interesting) by GreatAuntAnesthesia on Tuesday April 23 2019, @02:30PM (1 child)
So from the blurb above, it sounds like every Tesla already on the road has a socket somewhere for one of these chips. Pull into a dealership, get one plugged in and away you go.
Say what you like about Tesla and Musk, but offering that level of upgrade (even if it only fulfils half the promise) to customers who have already parted with their money is admirable. Can you imagine, say, a phone manufacturer doing the same? Most of them can't even be bothered to offer security updates because they see old models as a threat to newshiny sales.
And as to all the haters upthread... yeah, he oversells. No doubt. He says a lot of stuff he really probably shouldn't, and if I were his friend I would be quietly advising him to shut down his twitter account. But if in his sales pitches he he has confused what is possible today with what ought to be possible in the future, the result has been significant pressure on the incumbent manufacturers who, let's face it, would otherwise be quite happy to sit on their damn thrones and keep building shitty inefficient oversized petrol / diesel SUVs forever. Musk / Tesla have pushed the electric and self-driving markets hard. If Tesla went bust tomorrow, the world would be a better place for the changes it has already forced on other badges.
(Score: 2) by HiThere on Tuesday April 23 2019, @05:07PM
IIUC, it's only the recent models that have this upgrade capability. That said, it's interesting that it's already there.
Javascript is what you use to allow unknown third parties to run software you have no idea about on your computer.
(Score: 1, Interesting) by Anonymous Coward on Tuesday April 23 2019, @02:48PM
where i live we have scary roads.
the are ALIVE! on a 800 km strech you will encounter improvment, expansion etc work at least 5 times.
every trip is different. mostly it's getting better. in some rural areas alot of sugarcane is planted, harvested and sloppily
transported on the road. squishing sugarcane on asphalt doesn't seem to help asphalt live to a ripe old age, e.g. that stretch
gets upgraded/improved alot. mission: to make potholes extinct.
other areas are also pothole magnets, because they have a few houses in the middle of nowhere on both sides and the people
living there feel lonely and need some security at night, thus they feed dogs ... until they don't and the hungry dog starts roaming for antacid tablets, then
gets spilled with a tummy of hungry acid over the road ... not-vomited vomit doesn't mix well with asphalt.
note to self: don't squish a hungry dogonto the road. result will be instant pothole.
tho i am not sure if it's the acid or just that bones are harder then asphalt and if you keep rolling over them, they eventually generate a pothole?
anyways ... i hope the A.I. is also trained on section of so-called roads (in the making and upgrading etc).
protip: if it takes off, EVERYBODY will want it and the demand for A.I. assisting ROADPAINT will skyrocket ... abit.
in other news ... the tesla factory built itself ... on mars :)