Tesla, TSLA & the Investment World: the Perpetual Investors' Roundtable

[jerry33]

Pepsi(co) isn't using the Frito Lay semis to haul Doritos and such?

Considering the paint jobs I always figured they were double-dipping.

That's the soda news that'll be hard to wash down.

(we could keep chipping away at this...)

That's cola'n them like you see 'em. Let the chips fall where they may.

 

[Knightshade]

You know the APU of the infotainment system has like an old old Vega GPU with 11CU. Any games that require 16gb of ram is already a mismatch for the GPU of this chip. 5+ year old games can work decently on this chip but only at low resolutions (720p, low settings), which means 8gb of ram will be far from being saturated. I mean last gen high end dedicated GPU like a GTX 3070 only has 8gb of ram.

The 3070 besides being a couple years old wasn't ever "high end" even in that generation- it was the mid-range card- with the 3080, 3080ti, 3090, and 3090ti all coming above it (and all having more than 8GB of dedicated video memory, apart from the 8-16GB more most gamers would have for system memory). Plus even the 3070s 8GB was much faster memory than what you'd get with shared system memory and no dGPU.

All that said I think the fact they didn't include the dGPU in the 3/Y Ryzen cars that the S/X ones did was already handwriting on the wall we were never going to see a wide release of the higher end/steam type gaming stuff going mainstream local to the car.

To me, this makes way more sense as a cost saving than lumbar support or USS removal. Gaming is way down the list of features in a car.

... I mean, that's literally what point my original post was making...

Some more cost savings in HW4 infotainment

The savings going from 16 to 8GB of memory, and 128GB instead of 256GB of storage, is pretty small on a single car (maybe 10-20 bucks retail, so less at wholesale pricing)... but across a few million cars it'll add up.

 

[bruce4000]

Pepsi(co) isn't using the Frito Lay semis to haul Doritos and such?

Considering the paint jobs I always figured they were double-dipping.

That's the soda news that'll be hard to wash down.

(we could keep chipping away at this...)

PepsiCo had said that they haul chips around 250 miles, while drinks they have local bottling plants so only haul about 100 miles. Whether EV or Diesel, it’s better to ship the syrup and then add water locally.

 

[MP3Mike]

Pepsi(co) isn't using the Frito Lay semis to haul Doritos and such?

Again, the video being discussed is about the trucks at the Sacramento Pepsi facility. They have all kinds of details on the profile page: Run On Less: PepsiCo
(They haven't put up the profile for the Frito-Lay portion of the operation yet, but that one is based in New York.)

Pepsi is deploying Tesla Semis at its Sacramento, CA location in long-haul applications with heavy loads. Trip distances are up to 450 miles on a single charge.

For example about the loads/routes they are talking about:

They are hauling beverages, not chips.

Interesting charging information, 0-75% in 45 minutes. If it held 750kW the entire time that means 75% of the battery is ~550kWh. That would mean the battery is around 730kWh. (Which is close to them saying 450 miles of range and an average of 1.7kWh/mile=~765kWh battery.)

 

[Zaddy Daddy]

It sounds you assume Tesla's valuation is based on them running their own internal robotaxi service as opposed to people or companies buying Teslas and then allowing them to operate at robotaxis.

Doesn't really matter one way or another, as I'm talking about the technical hurdles needed to be achieved to get actual functioning robotaxis. So much is software based, and Tesla will be handling that. Whether or not a 3rd party purchases the vehicle and Tesla gets a % of revenue is a separate conversion.

They have to work on the "taxi" part of the product (e.g. finding place to pickup / dropoff) still even after FSD is safe enough to drive on roads. Waymo and Cruise are already at least facing this problem and working on solutions.

Interest rates impact affordability of all cars, not just Teslas. What Elon actually said



The total cost comparison between new cars can flip based on rates, deoending on if the Tesla is more expensive to start with and amount financed.

I don't care if other cars are less affordable too, either way Tesla makes less revenue and less profit, therefore lower valuation.

 

[Zaddy Daddy]

Hmmm.
Cruise and Waymo need LIDAR so their cars will never be as cheap as a Tesla robotaxi to build. The HD maps requirement means that they will only ever work in large populated urban areas with a ton of demand. Waymo are never sending an HD mapping car down my road, trust me.
And I have missed the announcement of the cruise/waymo supercharger roll-out, which means ironically they will probably be paying Tesla to charge their cars anyway.
Plus... the brand associated with the tech? Are you serious? Maybe in the 0.001% of the USA (one market) where they offer a service right now. I remember being in a heavy metal band that was very popular in a certain town in Devon, UK, but we were definitely not globally associated with the music.
As for capturing hearts and minds:
They cant work in fog:

Waymo robo-taxis struggle in thick San Francisco fog

Software fix coming – that'll be a braking change we'd actually like to see

www.theregister.com

They make people feel sick and stop in the wrong place:

Test riders in Waymo's self-driving cars won't stop complaining

About 10 percent of riders are displeased — or worse.

futurism.com

They anger local residents because of where they park:

Waymo woes: Tempe residents frustrated as idle self-driving cars park in their neighborhood

Waymo is expanding in the Phoenix area, but that is causing a level of headache for residents in one Tempe neighborhood, who say some of the self-driving cars are being parked on their streets when they are not picking up passengers.

www.fox10phoenix.com

They block traffic and colide:

U.S. safety regulators investigating GM's Cruise robotaxis blocking traffic, causing collisions

U.S. automotive safety officials have opened a formal probe into the autonomous driving system used by General Motors' self-driving vehicle unit Cruise.

www.cnbc.com

Yeah, I think I'll not be worried about waymo or cruise, even if they carry out a massive expansion soon to cover 0.002% of the US, I think I'll keep my Tesla stock.

Lidar and mapping costs are miniscule compared to the potential revenue brought in on a per car basis.

"it will only work in major urban areas" - well where do you think the majority of usage, revenue, and profits will be from?

You all realize that putting robotaxis in the top 30 urban metro areas is most of the revenue capture, right? Mapping is irrelevant.

Wow, relying on Tesla for supercharging...so? Tesla just stated they wouldn't even be charger other OEMs higher rates to use the network! This is not a big profit generator for Tesla.

Waymo and Cruise will have functional robotaxis in probably 20 metros areas within 2 years, and a larger presence in some of them (like SF) that will potentially hurt Uber and other competitors.

This casual thinking Tesla will just simply catch up because "reasons" is not sufficient dialogue.

 

[Zaddy Daddy]

Simply by ignoring FUD, like that which you so amply provide.

Another way would be to count the number of these vehicles the competitors have on the road currently growing this tremendous revenue and providing services, and then compare that to the number of FSD-ready Teslas that are already available to be Robotaxis. Perhaps you are unfamiliar with the term "orders of magnitude" and how this might be something worth taking into account.

Discussing Robotaxi competitors is not FUD. You all can't just take any critical angle as FUD, it's intellectually lazy. You know, just like saying there were any demand issues late last year would have probably gotten people banned.

Meanwhile orders plummeted and the stock dropped 75%, but hey at least TMC Investors forum didn't have to read anything from "concern trolls"!

P.S - Tesla HW3 will likely not be sufficient for robotaxis considering already high compute usage with much smaller models than what will be needed for high accuracy FSD - this is just the nature of machine learning. Aside from the fact that the camera resolution is mediocre and camera placement less than ideal. Either way, HW4 has a much better chance. On top of that, Tesla can't step function turn on a fleet of robotaxis, it will still be an S-curve ramp and so HW4 cars will be more than sufficient.

Cruise and Waymo are not going to have another “2-3 years of easy expansion”. They are going to burn billions on a model that can not scale. There is a reason Google cut Waymo loose.

No one is ready to deploy robotaxis st scale. Talk of Tesla robotaxis is premature as long as FSD has a single digit take rate.

Sorry Waymo is still part of Alphabet. The cash burn is truly a concern, but I don't see evidence for why structurally it cannot scale.

Software costs are enormous but are mosly fixed costs that will stay quasi-linear while revenues scale. Even remote operators won't scale linearly over time. People love to talk COGs reduction here, the same will certainly happen with Waymo and Cruise vehicles (moreso even).


Also:

I'm still not a troll. And the ratio of likes / dislikes / funny is very healthy!

 

[AMN]

Discussing Robotaxi competitors is not FUD. You all can't just take any critical angle as FUD, it's intellectually lazy. You know, just like saying there were any demand issues late last year would have probably gotten people banned.

Meanwhile orders plummeted and the stock dropped 75%, but hey at least TMC Investors forum didn't have to read anything from "concern trolls"!

P.S - Tesla HW3 will likely not be sufficient for robotaxis considering already high compute usage with much smaller models than what will be needed for high accuracy FSD - this is just the nature of machine learning. Aside from the fact that the camera resolution is mediocre and camera placement less than ideal. Either way, HW4 has a much better chance. On top of that, Tesla can't step function turn on a fleet of robotaxis, it will still be an S-curve ramp and so HW4 cars will be more than sufficient.




Sorry Waymo is still part of Alphabet. The cash burn is truly a concern, but I don't see evidence for why structurally it cannot scale.

Software costs are enormous but are mosly fixed costs that will stay quasi-linear while revenues scale. Even remote operators won't scale linearly over time. People love to talk COGs reduction here, the same will certainly happen with Waymo and Cruise vehicles (moreso even).


Also:

I'm still not a troll. And the ratio of likes / dislikes / funny is very healthy!

Sorry about your FOMO for not investing in TSLA 5 years ago but it's still not too late. It will be the largest company on Earth soon enough.

 

[MC3OZ]

They have to work on the "taxi" part of the product (e.g. finding place to pickup / dropoff) still even after FSD is safe enough to drive on roads. Waymo and Cruise are already at least facing this problem and working on solutions.

The Robotaxi fleet is likely to be driven by an app similar to Uber, any change in this area is a software / map update.

For pickup the customer will stand in a safe / convenient location, the default assumption for a Robotaxi fleet is that any safe pickup location is also a safe drop off location.

The next step is giving the customer the option to decline a drop off location and select a better one.

Initially some Tesla staff can just map safe pickup / drop off locations in cites.

In the long run FSD can try to learn the visual aspects of safe pickup / drop off locations.

IMO fundamentally Waymo and Cruise don't have a solution which economically scales to the required fleet size. In a fully mature system, Robotaxis are 10%-60% of the global vehicle fleet.

What would it cost Waymo and Cruise to put 1 Million Robotaxis on US roads, and how long would it take?

 

[Electroman]

Meanwhile orders plummeted and the stock dropped 75%

Huh? Orders plummeted? oh my God.. I am shaking now..

This hyperbole is approaching trollish levels. I was following you on the FSD comparison with Waymo, and then you exposed yourself on the 'demand is crashing thru the roof' boogey. And the stock crash had nothing to do with overall market crash of 22, and Elon selling shares to buy Twitter?

 

[Christine69420]

i really hope this is not going to be the case for MY and M3 5 year old vehicles I only have 49000 miles a rarely supercharge... if I was not a TSLA investor i would not have been so forgiving...

Good odds you'll be fine. Tesla needs to replace very few batteries.

 

[Thekiwi]

Lidar and mapping costs are miniscule compared to the potential revenue brought in on a per car basis.

"it will only work in major urban areas" - well where do you think the majority of usage, revenue, and profits will be from?

You all realize that putting robotaxis in the top 30 urban metro areas is most of the revenue capture, right? Mapping is irrelevant.

Wow, relying on Tesla for supercharging...so? Tesla just stated they wouldn't even be charger other OEMs higher rates to use the network! This is not a big profit generator for Tesla.

Waymo and Cruise will have functional robotaxis in probably 20 metros areas within 2 years, and a larger presence in some of them (like SF) that will potentially hurt Uber and other competitors.

This casual thinking Tesla will just simply catch up because "reasons" is not sufficient dialogue.

I did think the “The HD maps requirement means that they will only ever work in large populated urban areas with a ton of demand” was a bit of a strange criticism of waymo/cruise.

 

[Thekiwi]

I think it is fair to discuss the FSD competitors as a growing threat to TSLA valuation. Elon keeps insisting that a FSD Robotaxi fleet is the future cash cow of Tesla automotive, and yet we dont see Tesla operating a single driveleress robotaxi while competitors are now in 8 cities with rapid expansion plans underway.

There are two distinctly separate FSD robotaxi systems that are already operating successfully. One of those companies just said it only took a few weeks to tweak the system in preparation for adding Austin to its network.

(I’m glad they are entering Austin, it might annoy Elon enough seeing them every day that he actually pushes Tesla to start a driverless robotaxi trial).

I don’t care if these robotaxis aren’t going to operate in bumpkinville nowhereland with a population of 57 people - a robotaxi service will follow the Uber rollout and focus on where the vast bulk of demand is: cities.

Tesla may be ahead in developing a truly “drive anywhere“ FSD system, but that is a different product than the business model for a robotaxi service.

Elon is perhaps letting perfect (a true drive anywhere FSD system), get in the way of good (a city by city rollout of a driverless robotaxi fleet that requires hardcoding some intersections/maps)

 

[MC3OZ]

a robotaxi service will follow the Uber rollout and focus on where the vast bulk of demand is: cities.

Cities are big places, sometimes it takes 1-2 hours to drive from one side of a city to the other side. (Via the city centre, not a circular road.)

Some parts of cities are fringe areas with low populations, but some Robotaxi trips will want to go to those areas.

The additional considerations are:-

• Fleet size - convivence
• Cost per mile

If I am using a Robotaxi, I want one that always arrives fast and picks me up from wherever I am, then takes me to where I want to go, efficiently, and for a low price.

Sometimes in a city environment a competing service might beat a Tesla to the job and cost about the same amount, but if a Tesla is doing the same trip, I know the Tesla is making a higher profit margin.

IMO it will be very easy for Tesla to win any price war, most customers will not care about any premium aspects, but a clean car with comfortable seats, and good entertainment would be the expectation.

I don't want to pay more for a Robotaxi with a particular brand, and if I want to travel to the city fringe, I expect a Robotaxi to take me there.

 

[Singuy]

Lidar and mapping costs are miniscule compared to the potential revenue brought in on a per car basis.

"it will only work in major urban areas" - well where do you think the majority of usage, revenue, and profits will be from?

You all realize that putting robotaxis in the top 30 urban metro areas is most of the revenue capture, right? Mapping is irrelevant.

Wow, relying on Tesla for supercharging...so? Tesla just stated they wouldn't even be charger other OEMs higher rates to use the network! This is not a big profit generator for Tesla.

Waymo and Cruise will have functional robotaxis in probably 20 metros areas within 2 years, and a larger presence in some of them (like SF) that will potentially hurt Uber and other competitors.

This casual thinking Tesla will just simply catch up because "reasons" is not sufficient dialogue.

If you think -2.5 billion a year is miniscule. Psssst. I have a secret to tell you...the demand for robotaxi is not there...because it's not competitive to public transportation or owning your own car. The admission cost needs to reduce by 50% and waymo needs to get their destination 50% faster. Also there are multiple remote interventions per trip which is not exactly..robotaxi.

 

[SOULPEDL]

More Bullish on FSD and TSLA now. Why are we even debating this lol.

At one time, Waymo had a viable path forward because costs of the hardware would eventually come down with volume, and the driverless taxi model has great margins all else equal. $$$

Enter Tesla FSD, and there will be no comparison in ride cost and convenience. This will crush Waymo's business model overnight as the consumer will switch to cheapest avail (and faster, safer...). What's happening today with Ford and GM, ditto that with Waymo - again on margins (not to mention the other benefits mentioned).

OTOH, I don't think the masses are quite ready even if it were 10x safer. When friends ride along and I demo it, they are amazed and curious. However, if I ask if they would get in a robotaxi... most say "No way." They will adapt, S-style, convenience and cost driven. (I'm again reminded of my Mom and the new escalator situation as a child.)

For us, #1 use case would be to summons it to/from anywhere, no occupants. Numerous use cases there, could be most of the current FSD users as the value proposition is again raised another notch. FSD is not there yet, but I do like the DOJO compute curve ahead and the fact that FSD kept improving on the human supervised version of learning. So imagine the rate of progress when the computer chews on that data for a few more months.

Besides, Waymo reminds me of AGVs that follow lines on the floor in the factories and sometimes turn and branch, slow as ever. They can totally look straight ahead for pedestrians or obstacles and stop. Contrast this with someone driving around the factory in a golf cart. They are very different.

 

[nativewolf]

Cities are big places, sometimes it takes 1-2 hours to drive from one side of a city to the other side. (Via the city centre, not a circular road.)

Some parts of cities are fringe areas with low populations, but some Robotaxi trips will want to go to those areas.

The additional considerations are:-

• Fleet size - convivence
• Cost per mile

If I am using a Robotaxi, I want one that always arrives fast and picks me up from wherever I am, then takes me to where I want to go, efficiently, and for a low price.

Sometimes in a city environment a competing service might beat a Tesla to the job and cost about the same amount, but if a Tesla is doing the same trip, I know the Tesla is making a higher profit margin.

IMO it will be very easy for Tesla to win any price war, most customers will not care about any premium aspects, but a clean car with comfortable seats, and good entertainment would be the expectation.

I don't want to pay more for a Robotaxi with a particular brand, and if I want to travel to the city fringe, I expect a Robotaxi to take me there.

I would agree on this but I think entertainment is ...meaningless. Everyone will be looking at their phones/tablets/computers. Or chatting with friends bar hopping- the single most popular use of Uber.

 

[nativewolf]

More Bullish on FSD and TSLA now. Why are we even debating this lol.

At one time, Waymo had a viable path forward because costs of the hardware would eventually come down with volume, and the driverless taxi model has great margins all else equal. $$$

Enter Tesla FSD, and there will be no comparison in ride cost and convenience. This will crush Waymo's business model overnight as the consumer will switch to cheapest avail (and faster, safer...). What's happening today with Ford and GM, ditto that with Waymo - again on margins (not to mention the other benefits mentioned).

OTOH, I don't think the masses are quite ready even if it were 10x safer. When friends ride along and I demo it, they are amazed and curious. However, if I ask if they would get in a robotaxi... most say "No way." They will adapt, S-style, convenience and cost driven. (I'm again reminded of my Mom and the new escalator situation as a child.)

For us, #1 use case would be to summons it to/from anywhere, no occupants. Numerous use cases there, could be most of the current FSD users as the value proposition is again raised another notch. FSD is not there yet, but I do like the DOJO compute curve ahead and the fact that FSD kept improving on the human supervised version of learning. So imagine the rate of progress when the computer chews on that data for a few more months.

Besides, Waymo reminds me of AGVs that follow lines on the floor in the factories and sometimes turn and branch, slow as ever. They can totally look straight ahead for pedestrians or obstacles and stop. Contrast this with someone driving around the factory in a golf cart. They are very different.

Have you actually ridden in a Waymo?

 

[2daMoon]

Again, the video being discussed is about the trucks at the Sacramento Pepsi facility. They have all kinds of details on the profile page: Run On Less: PepsiCo
(They haven't put up the profile for the Frito-Lay portion of the operation yet, but that one is based in New York.)



For example about the loads/routes they are talking about:

View attachment 962588

View attachment 962589
They are hauling beverages, not chips.

Interesting charging information, 0-75% in 45 minutes. If it held 750kW the entire time that means 75% of the battery is ~550kWh. That would mean the battery is around 730kWh. (Which is close to them saying 450 miles of range and an average of 1.7kWh/mile=~765kWh battery.)

View attachment 962590

What about the driver?

Could they have had a bag of chips in the cab?

 

[Singuy]

Have you actually ridden in a Waymo?

https://twitter.com/x/status/1687657979158331393