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

[Webeevdrivers]

I got asked if the cybertruck was profitable in first quarter. Was that broken out in the 1st quarter report?

 

[Stretch2727]

I got asked if the cybertruck was profitable in first quarter. Was that broken out in the 1st quarter report?

They don't break out the profitability of different car lines. They have even clouded deliveries by mixing with S & X.

 

[kbM3]

FWIW I had one just an hour ago- exiting highway on the exit ramp (so unclear if it was on the v12 stack by then or not)- same as one I had a week ago in the same situation (but totally different exit ramp) where it seemed to think the shoulder was a lane and began diving into it...

Had another NON critical pure highway one (so still 11 presumably) where it needed to take an exit the traffic was backed up like half a mile to exit, it kept blissfully passing all the slowed/stopped cars 1 lane over, and then realizing just a couple hundred feet before the exit still doing 65 it had no chance to merge in I took over and managed to greatly annoy some folks getting in right at the exit ramp.

Only pure-city one I've had this week was when it got to a road closed no-thru traffic sign and just sat there with no idea what to do- so that was critical in the sense of wanting to not sit there forever, but not safety critical.

Disengagement categorization is very important. It doesn’t sound like your interventions (serious though they may be) were necessary to avoid an accident.

I do wish we could get some information from Tesla on how they categorize disengagements (e.g. accidental, imminent accident avoidance, car too rude, car not aggressive enough…) and what their progress looks like for every type. I can’t think of any information more important to investors at this time, but I’m sure there are reasons Tesla is not ready to share. For instance they may release software that they know will cause more disengagements of a certain type (e.g. the car is behaving rudely) in order to gather more data, or for some other reason. So it might be illogical to expect them to make steady progress in every single disengagement category every release. And thus be subject to misinterpretation by Wall Street.

 

[Thumper]

For the weekend, some Tesla history.

https://twitter.com/x/status/1784676257835180484

I took delivery of one of those cars. VIN 2383 on Jan 2, 2013. It was delivered directly to my house. Happy to have pitched in, although at the time it was a speculative buy. I thought for sure I would need to rent an ICE periodically, but never did need to. There were no superchargers yet. Fun time.

 

[SPadival]

I haven't seen anyone mention Tesla doesn't have enough compute in the US for US AI, hence they are spending billions to build it out. How in the heck are they going to be able to process the exported data out of China even if it was allowed?

Tesla is building & commissioning new data centres right now. Elon mentioned the spend will be $10 billion

H100s are not the only game in town. Tesla has proven (as with the car chip shortage) chip shortages don't slow them down. They just write the software for different chips. There are Chinese GPUs but they are on a US blacklist so we never hear about them. I don't know if that means an American company like Tesla can't use them in China or those chips simply can't be imported, I believe it's the later.

I am not an AI expert, but I don’t think it is a straight forward process to port over training to a new chipset. So no harm in Elon earning his non-salary as a skilled negotiator and practice the lost art of diplomacy between the two countries (something that US government seems incapable of)

 

[MC3OZ]

And the business in Asia is particularly risky margin wise. I think you'll see energy margins go even further south than auto, ie negative. People will be buying market share and Tesla will be competing with sellers that produce the battery cells Tesla needs. Wrapping cells into packs, throwing on a cabinet and bms, not rocket science.

The Hornsdale battery project started in 2017 at the time it was hand built and Tesla needed to source cells from Samsung.

At the time I can remember many claiming that the whole idea of big batteries was an extremely dumb idea and it would never work.

We are only now getting to the point where Megapack is a fully dialled product, with mature demand, and sufficient cell supply.

Power electronics and software are the other important parts of the product.

Software not only determines what the product can do, it is also a way of optimising income from a very versatile well performing product.

What I am talking about here is maintaining "grid stability" via FCAS, frequency and voltage inertia, replacing more expensive gas providers of those services with a cheaper service which provides are more accurate much more timely service. In turn many devices plugged into the grid are protected from power surges, over and under voltage events which damage a lot of equipment. And even more important is the fact that this helps keep the gird up and running.

Batteries don't just do all of this out of the box, Tesla had to write the software to make it happen, and the battery had to be engineered in a way which supported those services. An important part of creating a market for a product is showing what the product can do, and how it can save money or make money,

Yes, others can now copy and replicate these services, even if their software isn't as good..

But before Tesla did all of this, many claimed it couldn't be done.

2019-2023 Tesla scaled vehicle manufacturing, building 3 new factories, paying down debt and building up a healthily cash buffer, while doing that they developed Megapack and developed the market for Megapack.

Margins are probably better for energy storage now and that probably will be the case for 2-3 years, but that may flip in future.

Software and AI in part determine the margin on vehicles and energy storage batteries. What the cells can do is partially determined by software and hence software plays some role in determining the product margin.

For EVs I think margins will improve when interest rates drop, long term ICE margins are not a useful guide, because dealerships, parts, and service are not a significant part of the EV solution. As with energy storage, software is one factor helping Tesla margins.

 

[MC3OZ]

Tesla is building & commissioning new data centres right now. Elon mentioned the spend will be $10 billion



I am not an AI expert, but I don’t think it is a straight forward process to port over training to a new chipset. So no harm in Elon earning his non-salary as a skilled negotiator and practice the lost art of diplomacy between the two countries (something that US government seems incapable of)

I think all of the AI training software runs in Python and Tesla has developed a Python compiler / interpreter for Dojo.

 

[jerry33]

I took delivery of one of those cars. VIN 2383 on Jan 2, 2013. It was delivered directly to my house. Happy to have pitched in, although at the time it was a speculative buy. I thought for sure I would need to rent an ICE periodically, but never did need to. There were no superchargers yet. Fun time.

I had one too. VIN 3150

 

[MP3Mike]

Also of note-- apparently only AMD-based MCU cars are getting this, not Intel ones.

This is at least the 2nd significant feature divergence recently where older MCU vehicles even that paid for FSD features (with no upgrade path) are being left behind

From what I have seen Intel based vehicles are getting the new enhanced Tesla Vision Autopark, what they aren't getting is the new high-fidelity Park Assist.

 

[GatorMeat]

Well, Martin Viecha is still posting about Tesla on X, and he still has a Tesla badge on his X account. Maybe that promotion to CFO has given him fresh motivation?

https://twitter.com/x/status/1784282908452327803

This shows that Baglino and the last CFO didn't resign on their own, as they left the building immediately and didn't pre-anounce their 'retirement'

 

[Webeevdrivers]

They don't break out the profitability of different car lines. They have even clouded deliveries by mixing with S & X.

Thank you.

 

[STARR X]

I took delivery of one of those cars. VIN 2383 on Jan 2, 2013. It was delivered directly to my house. Happy to have pitched in, although at the time it was a speculative buy. I thought for sure I would need to rent an ICE periodically, but never did need to. There were no superchargers yet. Fun time.

I got mine too! January 21, 2013. VIN# 3101. Delivered to my work as there were no open service centers in Georgia yet.

 

[M|S|M MYP]

I do not really pay attention to the 24 hour market but if you are curious, check out what the $TSLA SP is doing currently

 

[Drumheller]

I do not really pay attention to the 24 hour market but if you are curious, check out what the $TSLA SP is doing currently

Or you could just post a number or screenshot instead of making others go look it up. Thanks for your help.

 

[Eugene Ash]

NO, we are not. When discussing project accounting rules differ. In most casts costs are obviously reductions in cash, but in many cases the costs and revenues are allocated based on proportionate activity. The financial statements will record 'unsatisfied contract obligations' or some such, and that will reduce as work has been performed. Cash flow may be entirely different with both advance payments and deferred payments also recorded. Those complexities are why project finance includes accountants, finance and technical performance people. Generalizing about these is fraught...

But there are rough indications which are those summarized by CERN as shown by @Musskiah.
Despite the complexities those financial statement excepts are 'close enough for government work'.

Just remember the other fast growing items that deal with Power Supply, including commercial ones and retail ones including Superhchargers and large supplier agreements for use of Autobidder with residential or commercial power supply and use. None of those are explicitly separated, thus far.

Those all seem to be on the verge of being material.

Because of growing share of stationary storage, I think it is worth floating and upvoting a question on Say.com for the next earnings call. e.g. "As investors seek to understand project accounting methods for Megapack projects: what are the yearly revenue and cashflow contribution that you expect, and the foreseeable committed ramp?"

It would be great to get clarity as to which mix of project accounting methods they use (milestone/calendar/cost/ completion) and how they recognize a) costs b) revenue and c) how lumpy are the cash inflows. These are multiyear projects and depending on the methods used, we might be noticing just the top of the sea Volcano.

 

[CatB]

I had one too. VIN 3150

Me three, March delivery. Two weeks earlier and I could have been part of the Broder-debunking Roadtrip - good reminder of the progress.
Spent the time waiting for delivery reading the Tesla.com forum, then somebody there suggested this one, and someone here suggested buying stock, today’s cost basis $2.52 - another good reminder of progress.

 

[dhanson865]

Or you could just post a number or screenshot instead of making others go look it up. Thanks for your help.

Sure, easy to do.

 

[M|S|M MYP]

Or you could just post a number or screenshot instead of making others go look it up. Thanks for your help.

Not everyone is curious

 

[geometro]

Sounds like FSD is evolving nicely but leaving those with 2024.8.x with nothing to experience… hoping that’s changing soon.

 

[Opus BC fan]

I got mine too! January 21, 2013. VIN# 3101. Delivered to my work as there were no open service centers in Georgia yet.

I got my P85 v#2449 on 12/28/2012...