123 kWh battery pack on Cybertruck (reported)

[Diehard]

Embarrassing. Zero reason not to put a larger pack

I wouldn’t say zero. Larger pack means ;

• less acceleration for the same power train
• less payload for the same suspension/tires/….
• More cells means greater chance of pack failure. greater cost for Tesla to do warranty replacement and greater cost for owner/insurance company to pay for out of warranty replacement.
• It means higher price or smaller margin.

Sponsored

 

[GhostAndSkater]

and yet, when 4680 Model Y was revealed, turns out it wasn't a long range, and not even the longest range, Model Y made available

instead, it was shorter range, and turned out to be used to make an affordable Model Y

set aside that they built only a few handfuls of them



regardless, the point isn't to has out the historical details

the point is that the historical details are FULL of reasons both for AND against just how numerous, how quickly, 4680 cells will be - much less to what end Tesla will deploy them (eg, people just assume they'll be used to make longer range vehicles, when Tesla has already shown once that's not necessarily their intended purpose)



but hey, i get it, from within the xwitter influenza echochamber of "recent tweet by Musk shows definitively that 4680 ramp is already to 10M cells/day," makes it hard to stay in critical thinking mode

4680 Model Y was made with V1 cells, 86.5 Wh and 828 cells for a 71.6 kWh pack

Second gen cells are 95 Wh, so the pack with the same config is 78.6 kWh. Currently 2170 packs are 82 kWh for US Model Y, 78 kWh for CN and EU Model 3/Y

Run rate is at around 5 GWh/y, which is 144k cells a day, enough from somewhere between 75 to 110 Cybertrucks a day, don't think they will hit that anytime soon, so cell supply isn't a constrain

And this data is not from a X post, it's from teardowns, lab analysis of cells and earnings calls remarks

 

[cvalue13]

between 75 to 110 Cybertrucks a day

even on your math, that’s a grand total of 23,500 to 33,000 CyberTruck’s a year, and not one Semi or Model Y

if in Q1, 8,000 - 9,000 trucks

so maybe we’re just talking past one-another regarding what ‘constrained’ means

but to be clear, I don’t for a moment intend to suggest batteries are the only constraint to early ramp of Cybertruck.

 

[ldjessee]

even on your math, that’s a grand total of 23,500 to 33,000 CyberTruck’s a year, and not one Semi or Model Y

if in Q1, 8,000 - 9,000 trucks

so maybe we’re just talking past one-another regarding what ‘constrained’ means

but to be clear, I don’t for a moment intend to suggest batteries are the only constraint to early ramp of Cybertruck.

Are you taking the position that the current rate will not increase for a whole year?!?!
Because your math seems to indicate that, unless I missed something.

 

[cvalue13]

Are you taking the position that the current rate will not increase for a whole year?!?!
Because your math seems to indicate that, unless I missed something.

no, we’re all talking in crayon here

other guy was suggesting 4680 ramp will be so sufficient for CT, MY, and Semi, they won’t know what to do with them all

and as evidence, he matched that they currently have enough cells to build *only* 3,000 trucks a month (and zero Y/Semi)

they’ll have to 7X that within the next 18mo, just to have enough cells for the 250K/yr run rate Tesla says is 18mo away

will Tesla do that? Sure, maybe (though I don’t take Tesla’s timing estimates very serious)

but that still doesn’t leave any for MY or Semi, much less the extra cells Tesla ‘won’t know what to do with’

 

[Drummerm]

Basically waiting for reveal at this point.

holding out for my most important feature. . . . Ventilated seats.

to be honest, ventilated seats are a very critical feature.

 

[ED_SFO]

no, we’re all talking in crayon here

other guy was suggesting 4680 ramp will be so sufficient for CT, MY, and Semi, they won’t know what to do with them all

and as evidence, he matched that they currently have enough cells to build *only* 3,000 trucks a month (and zero Y/Semi)

they’ll have to 7X that within the next 18mo, just to have enough cells for the 250K/yr run rate Tesla says is 18mo away

will Tesla do that? Sure, maybe (though I don’t take Tesla’s timing estimates very serious)

but that still doesn’t leave any for MY or Semi, much less the extra cells Tesla ‘won’t know what to do with’

In 12-18 months there will be several of Tesla's battery partners making 4680 cells in the US too. Tesla does not need to make Cybertrucks in a hurry, as it's not a projected to be a high volume vehicle like the Model Y, but they do need to sell enough for the investment they already put in. Sell enough with a decent margin so it won't weigh down other parts of the business.

 

[ldjessee]

My understanding is that 4680 line 1 is running, but not yet at full rate.
2-4 lines are installed or will be finished soonish, then they will ran making Cybercell (ie, 4680 gen2).
Then hopefully, by the time the experimental line has gen3 of the 4680 chemistry worked out, lines 5-8 can be built with that chemistry in mind.
We assume that these will be higher density cells (and other features mentioned at Battery Day; but maybe no 100% of those goals, but approaching).
What does that mean?
I am not going to throw a number at it, but I am going to guess that unless gen 3 is so awesome that the lines for making 4680 in Nevada start with that chemistry, they will probably keep the Semi on 2170s (or 18650s, not remembering which cell the semi uses).

I also expect either silicon or Lithium Iron (phosphate or not) versions of the 4680s will be made and if I had to point a finger, it would be Nevada for storage products and the lower trim mass production cars (3, Y, and Model2... maybe a 2WD/entry model Cybertruck?).

I tend to be more optimistic than pessimistic. Things can always go wrong, but if you are going to wish for anything, why not wish on the positive end of the stick (while still preparing, just in case, for a more negative result).

Only time will tell.

 

[GhostAndSkater]

even on your math, that’s a grand total of 23,500 to 33,000 CyberTruck’s a year, and not one Semi or Model Y

if in Q1, 8,000 - 9,000 trucks

so maybe we’re just talking past one-another regarding what ‘constrained’ means

but to be clear, I don’t for a moment intend to suggest batteries are the only constraint to early ramp of Cybertruck.

The thing is, we have a bit of information that tells us that 4680 production is about to grow rapidly, faster than Cybertruck ramp

Tesla spent all this time on just one 4680 line in Texas, ironing the kinks, now they go to a point that they are confident enough that they are ramping line 2 with all the improvements learned and modified from line 1, meaning the don't foresee big changes and problems anymore, and also starting lines 3 and 4

You don't put time an effort doing 4 of the same if you foresee you will need to modify a lot of equipment and processes, that is 4 times the work

Tesla will have 200 GWh worth of 4680 production installed on Texas be the end of next year (their words) and hopefully ramped up fully by the end of 2025, plus another 100 GWh on Nevada, but who know when

This year Tesla will consume from all suppliers and to all product lines around 150 GWh, so they have in the pipeline, ignoring third party suppliers, plans to more than quadruple their GWh deployed to 450 GWh deployed (again, ignoring growth from third party suppliers)

Not Semi, not Model Y or whatever can consume all that much, for context

500k Model 3/Y LR - 41 GWh
50k Semi - 45 GWh
1 million compact vehicles - 41 GWh
250k LR Cybertruck - 45 GWh

 

[TheLastStarfighter]

Semi doesn't use 4680 cells.

 

[GhostAndSkater]

Semi doesn't use 4680 cells.

For now, because they didn't have enough when they started it's pilot run, Semi is the vehicle that most benefit from it and almost certainty will use when they enter series production

They are making a 4680 factory in Nevada, where Semi will be built in high volumes

Right now there is somewhere around 50 thousands 2170 cells for it's pack, 4680 bring that down to 10 thousands, significantly easier, faster and cheaper manufacturing iwse

 

[cvalue13]

The thing is, we have a bit of information that tells us that 4680 production is about to grow rapidly, faster than Cybertruck ramp

Tesla spent all this time on just one 4680 line in Texas, ironing the kinks, now they go to a point that they are confident enough that they are ramping line 2 with all the improvements learned and modified from line 1, meaning the don't foresee big changes and problems anymore, and also starting lines 3 and 4

You don't put time an effort doing 4 of the same if you foresee you will need to modify a lot of equipment and processes, that is 4 times the work

Tesla will have 200 GWh worth of 4680 production installed on Texas be the end of next year (their words) and hopefully ramped up fully by the end of 2025, plus another 100 GWh on Nevada, but who know when

This year Tesla will consume from all suppliers and to all product lines around 150 GWh, so they have in the pipeline, ignoring third party suppliers, plans to more than quadruple their GWh deployed to 450 GWh deployed (again, ignoring growth from third party suppliers)

Not Semi, not Model Y or whatever can consume all that much, for context

500k Model 3/Y LR - 41 GWh
50k Semi - 45 GWh
1 million compact vehicles - 41 GWh
250k LR Cybertruck - 45 GWh

Ok so yes, we're talking past one-another.

You're talking about 4680 production lines installed by end of 2024 and production ramp completing by end of 2025 (setting aside you're relying on timing expectations from Tesla).

Which amounts to asserting there'll be plenty of 4680 by 2026, but that prior to that cells will be ramping - then just asserting/assuming that the cell ramp itself will outpace the Cybertruck ramp, by some uknown definition/assumption.

But it's completely reasonable to wonder or assume if Tesla's concurrent timing assertions about Cybertruck ramp (e.g., 125K line capacity in 2024, and reaching a production run rate of 250K/yr sometime in 2025), isn't merely tracking Tesla's underlying assumptions about the batter ramp you conceed will be occuring concurrently.

Now, in the near term, if not all of 2024, no doubt batteries wouldn't be the *only* thing constraining Cybertruck production run rates. They have a *log* of sh*t to figure out, dial in, and source (batteries are not the only constraining part).


But I don't understand how you not only believe that battery ramp will outpace Cybertruck ramp independently, but more to the point, how you can fault or assert (from nothing?) that people who don't believe that are somehow unfamiliar with some math.



What Tesla is doing in 2026 is really irrelevant to me at the moment (and I suspect many others). The cybertrucks being released this week, are likelly to be very different in material ways from any cybertruck being offered for sale in 2026. And, how many are being sold in 2026 depends entirely on how ramp goes in 24/25

 

[GhostAndSkater]

Ok so yes, we're talking past one-another.

You're talking about 4680 production lines installed by end of 2024 and production ramp completing by end of 2025 (setting aside you're relying on timing expectations from Tesla).

Which amounts to asserting there'll be plenty of 4680 by 2026, but that prior to that cells will be ramping - then just asserting/assuming that the cell ramp itself will outpace the Cybertruck ramp, by some uknown definition/assumption.

But it's completely reasonable to wonder or assume if Tesla's concurrent timing assertions about Cybertruck ramp (e.g., 125K line capacity in 2024, and reaching a production run rate of 250K/yr sometime in 2025), isn't merely tracking Tesla's underlying assumptions about the batter ramp you conceed will be occuring concurrently.

Now, in the near term, if not all of 2024, no doubt batteries wouldn't be the *only* thing constraining Cybertruck production run rates. They have a *log* of sh*t to figure out, dial in, and source (batteries are not the only constraining part).


But I don't understand how you not only believe that battery ramp will outpace Cybertruck ramp independently, but more to the point, how you can fault or assert (from nothing?) that people who don't believe that are somehow unfamiliar with some math.



What Tesla is doing in 2026 is really irrelevant to me at the moment (and I suspect many others). The cybertrucks being released this week, are likelly to be very different in material ways from any cybertruck being offered for sale in 2026. And, how many are being sold in 2026 depends entirely on how ramp goes in 24/25

Well, the only source for that I have is the earnings call, which Drew said more than once that they planned and were on track to ramp 4680 production well ahead of Cybertruck ramp

I can't find the exact quotes, but if you look at the transcripts of this year earnings call you will find it

 

[cvalue13]

Well, the only source for that I have is the earnings call, which Drew said more than once that they planned and were on track to ramp 4680 production well ahead of Cybertruck ramp

I can't find the exact quotes, but if you look at the transcripts of this year earnings call you will find it

then between us it boils down to a value distinction

you placing more value than I on the objective truth (or predictive value) of Drew's comments


it may be a fault of mine, but I tend to view Tesla's timing and ramp predictions with more than a jaundiced eye.

though i admit that, in this specific case, it's a double correlated production of both battery and CT ramp, so if he's equally wrong about both then they may still track (even if the date endpoints move together outward)



still, worth re-emphasizing, that cells are definitely not the only CT ramp constraint, parts-wise

 

[RVAC]

Most info we have seen points out that Austin never made V1 cells

Such as?

We've seen Tesla tweet and post photos of 4680 production in Austin well before line #2 was installed.

Regardless of cell output the elephant in the room is the issue of energy density, as was discussed in another thread. I know you're firmly in the double stack camp but if you entertain the idea that there is only a single stack, along with a pack length constraint of ~92in, the most you could fit is anywhere between 133-142 kWh usable (141-150 kWh gross) by removing/reducing the side bolsters and going 4 modules 8-cell wide. So not enough for a 500 mile version until there are significant (~20-30%) improvements in energy density.

Sponsored