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

[FSDtester#1]

It will be interesting to see just how high of a percentage of the semi powertrain is shared with SEXY.

So is Tesla now SEMI S3XY?

I wanted the Model 2 to come out first so we could be 2 S3XY.

Said Right Said Fred.

 

[2daMoon]

Sam Alexander roasts Barron's claim that GM is Catching UP to Tesla...

Good, cheap entertainment for the weekend.

 

[wipster]

Today I picked up my very first Tesla, Shanghai built, ordered in September. Invested since 2015 but now - at long last - I'll know what you guys are talking about.

View attachment 881051

Congrats! Beautiful car, when I get one it will be that color... unless Tesla comes out with an even cooler blue that is...

 

[Oil4AsphaultOnly]

If the aim is to bring the vehicle to a halt, and thereby eliminate the source of heat, then provided it is done early enough then @dhanson865 is quite correct.

The problem is if one is trying to make progress descending the hill in a controlled fashion, as opposed to simply stopping. That is the situation I pose for consideration.

But @dhanson865 is also quite correct that the Tesla control system will know how close to the limit it can drive before getting too close to an out-of-control situation.

My point is that things can go from 'tolerable' to 'disastrous' exceedingly quickly.

I suspect that @dhanson865 and I have both been exposed to some of the same prior industrial experience in this area, and that the pool of engineers facing this set of challenges has widened significantly in the last decade.

Right, that was what I was thinking of in my "edit", because you had referred to "leaving at the top of the mountain with a full charge and load". That's why it was specifically heat dissipation with only the motor + cooling system and a constant downhill speed. Is it the same amount of energy to be dissipated to maintain the same speed, regardless of what the actual speed is? I imagine that at higher speeds, aero drag and rolling resistance would help keep the speed under control?

 

[Gigapress]

Nope. I believe I recognise that behaviour when I see it, and this isn't it.

The person in question has been saying things I would have said myself if he/she had not already said them. And he/she has historically been very fair with respect to the facts, and most polite in setting them out. There are limits to what Tesla has disclosed and the poster is quite legitimately calling attention to the known edges of the story. In this case there are indeed a few thousand lbs of payload astray (horrible imperial units) and that is about a tonne or so in round numbers. Which in turn is perhaps 10% of payload for the use-cases that weight-out. (Which are a sigificant though minority of the use cases). Whilst I appreciate everyone who is trying to back-calculate the exact numbers, the point is that the Tesla specification sheet for the Semi is not yet available. If it was we could do the maths transparently ourselves rather than trying to fill in the gaps in a fuzzy fashion. The poster is quite simply correct to point this out.

Similarly the poster is quite simply correct to point out that what we are seeing is low rate initial production (LRIP). They haven't used the term, but they are quite correct that this is not full-on production or even a ramp of a 'final' production facility. If it were, we would have that spec sheet. And we'd know if this was the location of the final facility - which we all think is more likely to be Austin than LRIP in Sparks. So design & mfg iteration will likely be continuing (for how long) ?? prior to freezing on a next-stage production facility, and presumably that call will be aligned with cell/etc availability constraints relaxing.

I'm very happy to see the Semi finally getting to customer's hands. And I'm very happy to see Tesla's cautious approach to all this. And I'm very happy to see useful comments from those posters that show where Tesla's stated facts end, and over-optimistic infilling by others (like us) takes over. Because that in turn helps me keep my own positive biases in check.

What I'm not at all happy to see is poor behaviour coming from people who should know better, seeking censorship so they can stay in their happy zone. Whatever their fancy title.

No. @Tony73 published a bunch of BS FUD with no legitimate arguments backing up any of it.

Assertions:

1. 1.7 kWh/mile is unproven and “speculation” until an “official spec sheet” comes out
2. We can’t be sure Tesla will actually achieve 500 miles within weight and cost limits
3. No factory exists for production at scale and Elon nevertheless was loose with language and misleadingly called this “start of production”

Math was shown, with physics calculations, deriving 1.7 kWh/mile based on Tesla’s claim of the drag coefficient and assuming typical coefficient of rolling resistance for a semi. Elon himself corroborated this claim yesterday. Yet @Troy73 proceeds to dismiss this as “speculation” until an “official” spec sheet is published and until we know the exact conditions under which 1.7 kWh was achieved. That’s pedantic at best, because Elon’s public comment literally is official because the CEO is by definition an officer of the corporation, because he said that Tesla is already doing it on routes in the real world, and because from day one Tesla in fact already has published official specs advertising “< 2 kWh/mile” so 1.7 kWh/mile is right within the approximate range implied by that.

Tesla has been testing the truck on public roads for at least five years. A recently as this past May, they stated that they expect it can carry at least as much max payload as a diesel truck. Tesla knew precisely the weight of the tractor when making that statement. To suggest that it’s “speculation” to believe the production version will meet that spec is to imply that Tesla somehow was either lying or that Tesla was wrong in the design and in the last six months unexpectedly needed to add a bunch of weight to the tractor yet they nevertheless proceeded with initial deliveries to paying customers and chose not to disclose this information to the public. That is technically possible, but extremely unlikely, because it would damage Tesla’s reputation and probably invite false advertisement lawsuits, and because the design was almost certainly close to being finalized in May because there’s a lead time for making the assembly line and setting up the supply chain and everything else. Continuing to post here about how it’s not a proven fact until someone other than Tesla demonstrates it is again pedantic at best. Obviously we have to take Tesla’s word for it, but it’s pretty reasonable to believe their claim unless given a compelling reason otherwise, and Tony never even attempted to provide one.

Going on and on about cost is also silly, because Tesla has literally hundreds of thousands of dollars of wiggle room from the initial list prices from 2017 to still make a strong business case for customers, and that’s true before the gigantic clean energy subsidies for the batteries and the commercial clean vehicle credits (roughly $80k total subsidy per truck sold in the US). This is basic arithmetic. Based on the claimed efficiency number, the long-range truck has a battery around 900 kWh. Even if the pack-level cost for 2170s is conservatively assumed to be $125/kWh before subsidy, that’s $113k for the battery. If we conservatively assume that the rest of the Tesla Semi costs the same as a diesel truck with maybe $10-15k conservatively factored out for not having a diesel powertrain, then that’s a roughly $100k upfront cost difference. Fuel savings alone will make up for that disadvantage in just a couple of years for the average truck duty cycle. The only significant question about cost is just how good the economics are, but it’s obviously superior to diesel on total cost of ownership by a wide margin.

Would not surprise me [if this week’s 500 mile run was the first time Tesla had ever done it with their Semi]. The specs were obviously “aspirational” when they were revealed in 2017

Translation: “Tesla fraudulently advertised a key spec for this product in 2017 without actually knowing if they could achieve it within cost and payload constraints, and instead of providing math and evidence to back up this claim I will rather state that this is obviously true so as to signal my confidence.”

Wherever it is we can at least agree that Tesla does not yet have a production facility to produce semi at scale

The program is not fishy.
Elon just chose to call what happened yesterday “start of production” when it obviously is not.
They are probably still a year out. …

Here’s another vaguely defined claim, again with no supporting rationale, followed up by a thinly veiled insinuation that Elon deliberately misled us in choosing to characterize the event as a party kicking off “start of production”.

What does “at scale” mean, and how does @Tony73 purport to know anything about the production capacity of the Sparks factory despite there being, to my knowledge, zero publicly available information about the production line?

And even if the Sparks line is more of a pilot line with a bigger factory coming later, perhaps as an addendum to Giga Texas, who cares? That’s just more pedantry, considering that Tesla already disclosed on the Q3 call that they’re targeting a 50k annual rate in 2024, and have never said large scale production will occur in the next few months. When I worked at Boeing, I saw the 777X line get built and watched low-rate initial production begin ramping. When Boeing held a celebration kicking off manufacturing of planes that weren’t just test units and initial customer deliveries, nobody came around saying, “Well actually, that’s not real start of production; it doesn’t count until you’re making at least one plane per day.” If the company is building certified, legal products and selling them to customers, then that’s start of production by any reasonable definition.

 

[SmokyPeat]

Right, that was what I was thinking of in my "edit", because you had referred to "leaving at the top of the mountain with a full charge and load". That's why it was specifically heat dissipation with only the motor + cooling system and a constant downhill speed. Is it the same amount of energy to be dissipated to maintain the same speed, regardless of what the actual speed is? I imagine that at higher speeds, aero drag and rolling resistance would help keep the speed under control?

The steepness of the grade is not surprisingly a major factor. I don't think you'd want to be going at speeds that are heavily influenced by aero down a steep grade with little or no regen braking available. Maybe an emergency parachute like drag cars use? Kidding

 

[petit_bateau]

Right, that was what I was thinking of in my "edit", because you had referred to "leaving at the top of the mountain with a full charge and load". That's why it was specifically heat dissipation with only the motor + cooling system and a constant downhill speed. Is it the same amount of energy to be dissipated to maintain the same speed, regardless of what the actual speed is? I imagine that at higher speeds, aero drag and rolling resistance would help keep the speed under control?

I've not sat down and done any sums for the vehicle use case that I raise as a concern for the Tesla Semi (indeed any BEV). I could do so, but I'm about to go down the pub !

(It is in fact a more benign use case than the comparable one for another application that I have tangled with in the past, hence my interest and knowledge).

To help visualise how tight one needs to keep control, it is worth recalling that the Curie temperature of neodymium magnets is only 300-400 deg C, and there are only a few kg of these magnets in the three motors (wild guess, say 45 kg across 3 motors). In contrast the whole loaded truck weighs approximately 37,000 kg. So there is a lot of energy to be managed in just a few kg, with not much temperature rise allowed before things get out of hand.

 

[FireMedic]

So is Tesla now SEMI S3XY?

I wanted the Model 2 to come out first so we could be 2 S3XY.

Said Right Said Fred.

To clarify the most expensive trolling in corporate history.

It’s SEXY CARS
(Cybertruck, Atv, Roadster, Semi)

 

[JRP3]

I'm not sure what problem you think they need to solve.

My question is what happens in a complete electronics failure, i.e. high voltage disconnect blows? I assume different connections for each motor/inverter to provide redundancy so that potentially 3 different disconnects would have to be triggered.

 

[AudubonB]

My question is what happens in a complete electronics failure, i.e. high voltage disconnect blows? I assume different connections for each motor/inverter to provide redundancy so that potentially 3 different disconnects would have to be triggered.

Have we any knowledge - or supposition - of any single or dual motor Tesla vehicle that has had regen disabled because of such failure? If so, what has been the failure rate, either per size of fleet, or per total fleet-miles/km?

 

[brantse]

Have we any knowledge - or supposition - of any single or dual motor Tesla vehicle that has had regen disabled because of such failure? If so, what has been the failure rate, either per size of fleet, or per total fleet-miles/km?

I lost regen in my Model S one time, when the drive unit failed. The cause was from water ingress. It caught me off guard, as the car felt like it started accelerating all the sudden, but was from lack of regen braking. All the while, the dash was lighting up with many alarms.

 

[growler23]

The steepness of the grade is not surprisingly a major factor. I don't think you'd want to be going at speeds that are heavily influenced by aero down a steep grade with little or no regen braking available. Maybe an emergency parachute like drag cars use? Kidding

Does anyone else think we may be making a "tempest in a teacup" out of this "what happens to Semi on large downhills"?
A truck driver already knows the exact route before starting the drive. Very likely they have driven it 50 times before. Therefore, they will know exactly how much to charge to leave empty due to an initial downhill, for example. And yes, in the unlikely event the truck is cold started, they will (gasp) use more friction braking at the start until everything warms up and full(er) regen can kick in. Heck, even I use "preconditioning" to make sure that scenario doesn't happen on cold days - I suspect every trucking company would do the same as a no brainer. Perhaps Tesla calls this megapreconditioning.
I really don't think we have an issue here.
If we are delivering icebergs from the top to the bottom of Mount Everest, I stand to be corrected and agree we can speculate about giant resistor banks, jury-rigged power transmitting Starlink dishes, or the like

 

[SmokyPeat]

Does anyone else think we may be making a "tempest in a teacup" out of this "what happens to Semi on large downhills"?
A truck driver already knows the exact route before starting the drive. Very likely they have driven it 50 times before. Therefore, they will know exactly how much to charge to leave empty due to an initial downhill, for example. And yes, in the unlikely event the truck is cold started, they will (gasp) use more friction braking at the start until everything warms up and full(er) regen can kick in. Heck, even I use "preconditioning" to make sure that scenario doesn't happen on cold days - I suspect every trucking company would do the same as a no brainer. Perhaps Tesla calls this megapreconditioning.
I really don't think we have an issue here.
If we are delivering icebergs from the top to the bottom of Mount Everest, I stand to be corrected and agree we can speculate about giant resistor banks, jury-rigged power transmitting Starlink dishes, or the like

It's an interesting discussion. But yes, I'm sure the engineers at Tesla have well thought out safety in the braking system. I wouldn't be at all surprised if dynamic resistor braking becomes an option at some point. Traditional diesel semis rely heavily on compression braking (Jake brakes) in situations that call for heavy and/or sustained braking so as not to overload the hydraulic friction brakes. Obviously that isn't an option for an electric semi, and regen braking can't be utilized at all times for the reasons mentioned. In the future there might well be overhead catenary or rail systems on particularly long/steep grades to allow pulling energy from the grid on ascent and feeding it back on descent when necessary.

 

[Unpilot]

No. @Tony73 published a bunch of BS FUD with no legitimate arguments backing up any of it. The forum is better off without clutter from that junk.

Assertions:

1. 1.7 kWh/mile is unproven and “speculation” until an “official spec sheet” comes out
2. We can’t be sure Tesla will actually achieve 500 miles within weight and cost limits
3. No factory exists for production at scale and Elon nevertheless was loose with language and misleadingly called this “start of production”

Math was shown, with physics calculations, deriving 1.7 kWh/mile based on Tesla’s claim of the drag coefficient and assuming typical coefficient of rolling resistance for a semi. Elon himself corroborated this claim yesterday. Yet @Troy73 proceeds to dismiss this as “speculation” until an “official” spec sheet is published and until we know the exact conditions under which 1.7 kWh was achieved. That’s pedantic at best, because Elon’s public comment literally is official because the CEO is by definition an officer of the corporation, because he said that Tesla is already doing it on routes in the real world, and because from day one Tesla in fact already has published official specs advertising “< 2 kWh/mile” so 1.7 kWh/mile is right within the approximate range implied by that.

Tesla has been testing the truck on public roads for at least five years. A recently as this past May, they stated that they expect it can carry at least as much max payload as a diesel truck. Tesla knew precisely the weight of the tractor when making that statement. To suggest that it’s “speculation” to believe the production version will meet that spec is to imply that Tesla somehow was either lying or that Tesla was wrong in the design and in the last six months unexpectedly needed to add a bunch of weight to the tractor yet they nevertheless proceeded with initial deliveries to paying customers and chose not to disclose this information to the public. That is technically possible, but extremely unlikely, because it would damage Tesla’s reputation and probably invite false advertisement lawsuits, and because the design was almost certainly close to being finalized in May because there’s a lead time for making the assembly line and setting up the supply chain and everything else. Continuing to post here about how it’s not a proven fact until someone other than Tesla demonstrates it is again pedantic at best. Obviously we have to take Tesla’s word for it, but it’s pretty reasonable to believe their claim unless given a compelling reason otherwise, and Tony never even attempted to provide one.

Going on and on about cost is also silly, because Tesla has literally hundreds of thousands of dollars of wiggle room from the initial list prices from 2017 to still make a strong business case for customers, and that’s true before the gigantic clean energy subsidies for the batteries and the commercial clean vehicle credits (roughly $80k total subsidy per truck sold in the US). This is basic arithmetic. Based on the claimed efficiency number, the long-range truck has a battery around 900 kWh. Even if the pack-level cost for 2170s is conservatively assumed to be $125/kWh before subsidy, that’s $113k for the battery. If we conservatively assume that the rest of the Tesla Semi costs the same as a diesel truck with maybe $10-15k conservatively factored out for not having a diesel powertrain, then that’s a roughly $100k upfront cost difference. Fuel savings alone will make up for that disadvantage in just a couple of years for the average truck duty cycle. The only significant question about cost is just how good the economics are, but it’s obviously superior to diesel on total cost of ownership by a wide margin.


Translation: “Tesla fraudulently advertised a key spec for this product in 2017 without actually knowing if they could achieve it within cost and payload constraints, and instead of providing math and evidence to back up this claim I will rather state that this is obviously true so as to signal my confidence.”



Here’s another vaguely defined claim, again with no supporting rationale, followed up by a thinly veiled insinuation that Elon deliberately misled us in choosing to characterize the event as a party kicking off “start of production”.

What does “at scale” mean, and how does @Tony73 purport to know anything about the production capacity of the Sparks factory despite there being, to my knowledge, zero publicly available information about the production line?

And even if the Sparks line is more of a pilot line with a bigger factory coming later, perhaps as an addendum to Giga Texas, who cares? That’s just more pedantry, considering that Tesla already disclosed on the Q3 call that they’re targeting a 50k annual rate in 2024, and have never said large scale production will occur in the next few months. When I worked at Boeing, I saw the 777X line get built and watched low-rate initial production begin ramping. When Boeing held a celebration kicking off manufacturing of planes that weren’t just test units and initial customer deliveries, nobody came around saying, “Well actually, that’s not real start of production; it doesn’t count until you’re making at least one plane per day.” If the company is building certified, legal products and selling them to customers, then that’s start of production by any reasonable definition.

This is so much better than what I wanted to write.
Which was..." they are full of something that does not smell nice and is generally brown"

 

[StarFoxisDown!]

Another way to boost margins: FSDbeta discounted to $9000 for SR RWD Model 3.

https://www.tesla.com/new/5YJ3254_7c68d222816520d153cf0004c6e80680

Link not loading for me.

But I actually think them doing a FSD discount to say $7000 from $12000 or 1 year free FSD subscription for any deliveries in December would be an insanely smart move. They wouldn't compromise current margins and I think that would drive a ton of demand for Dec. I'm actually not sure why they're not doing that sort of incentive instead of the temporary price cut

 

[ggr]

CA (Neuralink) on Wed., NV (Semi) on Thur., DC (Macron) on Fri.

Hope he's productive on that jet.

Not terribly relevant, but he met Macron in New Orleans, not DC.

 

[SmokyPeat]

Link not loading for me.

But I actually think them doing a FSD discount to say $7000 from $12000 or 1 year free FSD subscription for any deliveries in December would be an insanely smart move. They wouldn't compromise current margins and I think that would drive a ton of demand for Dec. I'm actually not sure why they're not doing that sort of incentive instead of the temporary price cut

Not a bad idea, but I'm sure they have a good handle on which demand levers to pull. I don't see how free FSD wouldn't affect margins, however.

 

[MP3Mike]

My question is what happens in a complete electronics failure, i.e. high voltage disconnect blows? I assume different connections for each motor/inverter to provide redundancy so that potentially 3 different disconnects would have to be triggered.

You stop using friction brakes and call a tow truck.

 

[mongo]

They don't have to put power back into the packs to brake. In fact they can brake harder if they pull power from a full pack.

How does drawing from the battery help? Negative torque (opposing motion) is regen.

Right, that was what I was thinking of in my "edit", because you had referred to "leaving at the top of the mountain with a full charge and load". That's why it was specifically heat dissipation with only the motor + cooling system and a constant downhill speed. Is it the same amount of energy to be dissipated to maintain the same speed, regardless of what the actual speed is? I imagine that at higher speeds, aero drag and rolling resistance would help keep the speed under control?

Higher speeds need less braking energy (kWh/mile) but more braking power (kW) to maintain a constant speed.
Fully loaded, a 1% down grade is about 1.7kWh/mile so at 60 MPH it roughly balances out (0.5% to rolling 0.5% to aero). 3% would need about 2 times the speed to balance (1% rolling 2% aero).

 

[wipster]

While I'm a big proponent of a catch all thread, it is amazing how the perpetual investors thread seems to have become the Semi thread this weekend...