Tesla Live Stream – Autonomy Day [video] • Hacker News Board

Re: Tesla Live Stream – Autonomy Day [video]

#531

Post by tomp » Tue, Apr 23, 2019, 11:27 PM UTC

adrianmonk wrote: post #450

p1esk wrote: post #360

Earlier quoted context omitted.

This is still unsolved research problem Luckily, Karpathy is pretty good at solving research problems. Karpathy could have easily spent entire day talking about current state of the art and how they have improved on it Why would he do that? To help Uber? Collecting data is actually more trivial part of this problem. For Tesla, yes. For all others that’s the hardest part.

When they say that collecting data is the trivial part, they mean that (in comparison to the other parts of the problem) it is easy for everyone to solve. In other words, they are saying Tesla has an advantage in area that doesn't matter.

It’s “easy” in the sense that reusable rockets are “easy” - they’re possible, but noone else is working on them.

Musk was way ahead of time, again.

Re: Tesla Live Stream – Autonomy Day [video]

#532

Post by leesec » Tue, Apr 23, 2019, 11:38 PM UTC

Fricken wrote: post #512

leesec wrote: post #358

Earlier quoted context omitted.

>"We've gone over this multiple times. Are we sure we have the right sensor suite? Are we sure we have everything we need?.... No." Where did he say this? He repeatedly said they have everything they need to drive. >but before you can deploy, you need to integrate Lidar, Why? >not all the conventional things you need to build a complete autonomous vehicle. No one knows how to make a completely autonomous vehicle beca…

There's a reason why nobody else in the industry thinks they can make robotaxis work without Lidar. Of course, if you were familiar with the rest of the industry you wouldn't say Tesla has gotten farther than anybody else in the field, that's just ignorant.

Well I'm a software engineer at an automotive tech company who interfaces with all the big OEM's daily so I'm gonna go ahead and say I'm very familiar. But that's not a reason you know, that's just you diverting.

Re: Tesla Live Stream – Autonomy Day [video]

#533

Post by millettjon » Wed, Apr 24, 2019, 12:36 AM UTC

b_tterc_p wrote: post #162

sairahul82 wrote: post #114

Karpathy's presentation is really good. Watch it later if you get some time. The key points are - Telsa is using fleet for learning (Fleet learning). This has 3 components, a. Trigger infrasture that collects the kind of data telsa is looking for training. The rational here is we don't need massive amounts of same kind of data but needs right kind of data for training neural networks correctly. b. Data Engine which l…

> The rational here is we don't need massive amounts of same kind of data but needs right kind of data for training neural networks correctly. This is really important. At a certain point additional miles on similar roads in nice conditions with fair traffic won’t help anymore. The data variety is really important.

They also probably want to learn from the best human drivers and not repeat the mistakes of the average.

Re: Tesla Live Stream – Autonomy Day [video]

#534

Post by millettjon » Wed, Apr 24, 2019, 12:58 AM UTC

gamblor956 wrote: post #318

dreaminvm wrote: post #243

Earlier quoted context omitted.

They cherry-pick rare cases and use their fleet to get more examples of these situations. This seems like the right approach given more miles following the same car in a straight line is pretty useless. My takeaway from the presentation is that Tesla will perform better than other companies in this space (although I don't know enough about Waymo to comment) due to the following: -You want a large dataset (Tesla and m…

My point was that it appears that Tesla doesn't have a large and varied dataset. It has a small and pre-selected data set, since the cars only transmit data when pre-determined triggers are fired. Thus, it doesn't matter how many 1000s of drivers Tesla "has" or how many "situations" they're in, since it's not actually collecting data from most of these situations. And Autpilot's performance (including its numerous re…

Shadow mode would also transmit data no?

Re: Tesla Live Stream – Autonomy Day [video]

#535

Post by bg24 » Wed, Apr 24, 2019, 4:24 AM UTC

I read through all comments. Very informative.

How many people and companies are going to freak out if Elon and his team deliver FSD without LIDAR? Say in 2022, not even 2020. My layman logic is that AI is improving exponentially, and we do not know what artificial vision can do in 2 years.

Re: Tesla Live Stream – Autonomy Day [video]

#536

Post by TheSpiceIsLife » Wed, Apr 24, 2019, 4:47 AM UTC

robszumski wrote: post #286

metafunctor wrote: post #207

Earlier quoted context omitted.

I think the point was that lidar is too low resolution. You must have very high resolution (in both time and space) to achieve L4 or L5; to accurately determine if a pedestrian is not paying attention, or if a bicyclist is looking to cross the road. Not that Tesla is doing any of that right now, they most likely are not. But a lidar is not going to, either. Not ever, which is the point. I thought the fleet learning a…

I haven't seen any mentions of cheaper Lidars, like single axis scanning only, to get some sort of depth map without spinning mirrors or 360 degree positions that ruin aesthetics. I feel like this would be able to detect trees in the distance and other objects to orient the road/lane detection. And keep it cheap. Possibly add on to existing cars.

Search HN for 'solid state lidar'

https://hn.algolia.com/?query=solid%20state%20lidar&sort=byP...

Re: Tesla Live Stream – Autonomy Day [video]

#537

Post by navigatesol » Wed, Apr 24, 2019, 12:06 PM UTC

DeonPenny wrote: post #374

navigatesol wrote: post #224

Earlier quoted context omitted.

> It could easily send way more data than any care company at any time including over WIFI. Except that it isn't, and even Karpathy said the quantity doesn't matter, it's the data quality.

What they are sending way more data because from our knowledge GM and Ford are sending back 0 data and Waymo doesn't have half a million cars worth of data internally to pick from.

>from our knowledge GM and Ford are sending back 0 data

Yes, two of the autonomous vehicle leaders are not using any data whatsoever.

I thought this was the smartest forum on the internet?

Re: Tesla Live Stream – Autonomy Day [video]

#538

Post by s_y_n_t_a_x » Wed, Apr 24, 2019, 2:09 PM UTC

greedo wrote: post #516

s_y_n_t_a_x wrote: post #432

Earlier quoted context omitted.

Shut down the company the trunk is slightly off. Tesla is new to the game, give them a few more cars to catch small stuff like that.

Tesla has been around for 16 years, so I think it's not too critical to expect them to be able to sweat the small stuff.

16 years is a tiny fraction compared to the other car companies you're comparing it to.

But ask yourself, what's more likely, the other car companies catch up to their AI and battery tech, or Tesla fixes the trunk in the next model?

Re: Tesla Live Stream – Autonomy Day [video]

#539

Post by dragontamer » Wed, Apr 24, 2019, 4:34 PM UTC

Robotbeat wrote: post #501

shaklee3 wrote: post #494

Earlier quoted context omitted.

Nvidia already responded: ""Tesla was inaccurate in comparing its Full Self Driving computer at 144 TOPS of processing with Nvidia Drive Xavier at 21 TOPS," a spokesman said in an email. "The correct comparison would have been against Nvidia's full self-driving computer, Nvidia Drive AGX Pegasus, which delivers 320 TOPS for AI perception, localization and path planning." The statement also contends that "while Xavier…

From what I understand, Pegasus consumes about 500Watts, compared to under 100 Watts for Tesla's FSD computer. Elon in particular emphasized the performance per watt (as it's always possible to cram more chips to increase performance if you ignore cost and power consumption). The comparison made in the video: 500Watts for an hour consumes about 2-3 miles of range. In a city in slow traffic, going 12mph, that's a sign…

Its incomparable though. The Pegasus has far more compute power.

Just because the Pegasus has 500W worst-case TDP doesn't mean that its average case would be 500W constant. If you scale back your code and idle parts of the GPU, you can drop the energy cost arbitrarily.

At least, that's how GPUs on desktops work. They only use a ton of power if you give them a ton of work. Write your code in an energy-efficient manner, and the 2080 Ti will drop down to 20W, or scale all the way up to 300W.

https://www.tomshardware.com/reviews/nvidia-geforce-rtx-2080...

Modern chips idle very well. With the right code, the Pegasus could be tuned to only use 100W (assuming good enough programmers). But Tesla's chip will NEVER be able to scale above 100W.

Re: Tesla Live Stream – Autonomy Day [video]

#540

Post by Robotbeat » Wed, Apr 24, 2019, 6:58 PM UTC

dragontamer wrote: post #539

Robotbeat wrote: post #501

Earlier quoted context omitted.

From what I understand, Pegasus consumes about 500Watts, compared to under 100 Watts for Tesla's FSD computer. Elon in particular emphasized the performance per watt (as it's always possible to cram more chips to increase performance if you ignore cost and power consumption). The comparison made in the video: 500Watts for an hour consumes about 2-3 miles of range. In a city in slow traffic, going 12mph, that's a sign…

Its incomparable though. The Pegasus has far more compute power. Just because the Pegasus has 500W worst-case TDP doesn't mean that its average case would be 500W constant. If you scale back your code and idle parts of the GPU, you can drop the energy cost arbitrarily. At least, that's how GPUs on desktops work. They only use a ton of power if you give them a ton of work. Write your code in an energy-efficient manner…

Pegasus's chip is more general purpose and doubtless has far more general purpose compute power. But that's irrelevant. What's relevant is Tesla's chip is optimized specifically for their NN pipeline whereas Pegasus is based off of general purpose GPU architecture, thus Tesla's chip achieves a better TOPS per Watt than Pegasus. And it'd be strange if it didn't.

"Tesla's chip will NEVER be able to scale above 100W" okay, based on what? Tesla has a higher performance chip in the pipeline right now, and they could've used more silicon to achieve more TOPS if they needed it.

EDIT: Pegasus has 500W at 320 TOPS. Tesla's has 72 Watts at 144TOPS. Thus Tesla's chip, because it's focuses specifically on Tesla's NN pipeline, is about to get almost 4 times to performance per watt of the Pegasus and is much cheaper. Tesla's NN chip wouldn't help your video game, and Tesla isn't intending to compete in all the markets Nvidia operates in.