[keithwarren]

Asserting that Tesla is behind because they are fab at 14 instead of something smaller is so laughable you cannot be serious.

Thermal profile and power, instruction set and a dozen other things are more important.

Watch the autonomy day video on how the chip was built and why, will help a lot

[dragontamer]

> Watch the autonomy day video on how the chip was built and why, will help a lot

I watched it, and I'm calling them out on it. They focused their entire presentation on NVidia and completely ignored Google/Waymo and Intel/MobilEye Q5 comparisons.

Autonomy day was an obvious hype-fest that ignored the competition. Tesla was only looking at NVidia Xavier or Pegasus, but ignores the real giants in the field.

The arguments in that presentation were exceptionally poor. There's no way a 500W chip could be a problem in a car with 300,000W motors attached and 1000W air conditioners / heaters. It was obvious to me when they were skewing their examples to be absurd corner cases to make a point: 12mph stop-and-go traffic for many hours?

Yeah, as if the 300,000W motor would like 12mph stop-and-go traffic anyway. Its not like regenerative brakes are much better than 50% recycling of the energy btw, the stop-and-go condition is going to wreak your mileage regardless.

A 500W chip can be powered by a 75kW-hr battery for 150 hours. That's nearly a whole week of life. The thought that a Model 3 or Model S car would be scrounging for efficiency in the ~100W magnitude is completely laughable.

[matz1]

Vertical integration. Tesla can have the chip solely focused on their specific needs.

They ignore google/waymo because they won't sell them the chip anyway.

[dragontamer]

Elon Musk is pretending like Tesla can make money from this self-driving taxi idea.

You can't make money from self-driving Taxis if Google/Waymo simply offers a similar service for cheaper right next to you. And yes, Google/Waymo has the vertical integration thing all figured out too.

Now if you think that maybe Tesla's chip is useful for its M3 customers... sure. I strongly disagree with that fact, but you're certainly entitled to your opinion on that subject. My main issue is that Elon Musk is trying to sell this idea... that Tesla can actually raise money by going into self-driving taxis.

That's completely hogwash. Completely. There's not a chance of that at all. The other companies are too far ahead, and Tesla doesn't have enough cash to enter that fledgling industry.

Tesla isn't a tech company anymore, its a car company. It built a $2.5 Billion factory that can only produce cars, and its stuck with that (and its associated loans). Elon Musk can't just pivot into a new industry with all of these loans weighing down on his company.

Now, maybe if Elon Musk wanted to start a new company, one that wasn't burdened by all of the Gigafactory Debt, to tackle the self-driving problem... well... maybe that would work. But Tesla has already bet its life on wide-scale deployment of the Model 3. It doesn't have any other chances. It simply doesn't have the money to do anything else, or to pivot anymore.

[matz1]

>if Google/Waymo simply offers a similar service for cheaper right next to you

Right, if, tesla currently has the advantage being ahead, they have self driving (at some level) currently deployed for masses.

[dragontamer]

And Waymo has self-driving, with 10,000 miles per 2 or 3 disengagements... deployed right now in Phoenix Arizona.

Safety drivers are still necessary for both. But the number of times the safety-driver interacts in Waymo/Google's car is way way less.

EDIT: Actually, I'm not sure about Tesla's numbers. I heard 1 per 3 miles but I couldn't prove it through a google search. In any case, Tesla doesn't seem to be bragging about its disengagement rate, so I bet it isn't very good.

[matz1]

And tesla have like maybe 100000+ car with self driving in the wild right now. This also provide tesla with tons of real live data.

[dwighttk]

>Elon Musk is trying to sell this idea... that Tesla can actually raise money by going into self-driving taxis.

While giving as much as $30K/year of the revenue to each of the car owners.

[hef19898]

Vertical integration is dead in the automotive sector for decades. There are Ford Motor Companies of old anymore with their proper plantation to get the rubber for their proper tyres. And there are reasons for that, hard learned ones. Not sure why that would be any different for chips.

First, the likes of nVidia have a much larger customer base and thus a better risk profile. Also, chips are their core competency. And second, running your own chip production is capital intensive, not so great when you are already in a tight financial situation.

[Robotbeat]

Frankly, it'd be insane to invest in 10 or 7 or 5nm at this point for a custom design of maybe a few million units. The investment required scales insanely with reduction in feature size (Billions of dollars, potentially), and you'll never get your investment back. That effectively locks you in to an old design for several years. And the difference in unit costs or efficiency between 14 and 10nm is not big enough to matter compared to the upfront costs.

Fabbing at 14nm is the only sensible thing for Tesla to do right now, and since they're able to do it to replace their existing Nvidia chip for lower cost, it seems quite sensible.

[dragontamer]

Intel / MobilEye is making a 7nm chip for all of its customers. And part of that is because MobilEye has Ford, GM, BMW, Audi...

MobilEye is deployed in 313 car models across 27 manufacturers. They're going to have a 7nm chip next year, and are easily going to leapfrog Tesla's 14nm capabilities.

[Robotbeat]

If it's deploying so broadly over so many different models and manufacturers (with different sensor configurations, different IP, etc), then it necessarily has to be less application-specific and more general purpose. Tesla introduced its NN chip already in 2019 in actual built Model 3s already.

MobileEye plans to maybe have a 7nm chip sometime in 2020 (probably the end of the year) which might make it into year 2021 cars. So again, we're talking about ~2 years difference in deployment time for a less-tailored chip with a far more expensive development process. Maybe MobilEye can justify that (I'm not saying MobilEye is wrong to pursue 7nm), but it seems pretty clear Tesla pursued a logical plan to go with a 14nm process that is much cheaper to design for and is already available now at high quantities from multiple providers and reliable yields with zero technical risk on the chip fab side.

[dragontamer]

ASICs are all well and good when you can afford them AND when you have the volume to sustain them.

The main issue is that Tesla can't afford them. They have $2.2 Billion in cash remaining and are currently losing money at a rate of $700 Million/quarter.

Tesla's cash reserves are nose-diving and they're busy doing vanity projects like a self-driving ASIC, when their volume of cars being produced is well under 300k / year.

The volume simply isn't there yet to sustain this ASIC. 300k/year total car production, and not all of those cars would even have a self-driving capabilities installed. Tesla needs to work on expanding its Model 3 capabilities before ASICs make sense.

70,000 total vehicles produced Q1 2019. That's 280,000 vehicles this year if extrapolated over 4 quarters.

[Robotbeat]

All the more reason for them to go with the proven and far cheaper 14nm process node rather than 7nm (which Intel/MobilEye have had difficulties with as far as yields). https://www.techpowerup.com/248008/intel-at-least-5-years-be...

EDIT: Corrected below. MobilEye is using TSMC even though they're owned by Intel.

[dragontamer]

Intel/MobilEye is using TSMC 7nm for the EyeQ5 chip. There won't be any issues.

https://www.eetimes.com/document.asp?doc_id=1333990

TSMC 7nm has already deployed the iPhone A12, the iPad A12x, and AMD's Radeon VII. Its mature and ready for mass production.

-------

The problem with Tesla's 14nm chip is that their capabilities are about to be leap-frogged by a commercial-off-the-shelf solution in just a year. Note that the EyeQ5 is sampling TODAY. MobilEye is aiming at cars being deployed with the thing for 2020.

[hef19898]

More like all the more reason to source these chips from outside, I would argue