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

[mongo]

Perzactly. And the traditional manufacturing method for body panels is stamping. Can't do that with SS, so it's not just a matter of it being harder to do the traditional way and hence needing to be careful, but rather entirely new production methods. The initial prototype needing to be laser cut before bent/folded. Not sure what the ultimate mass production approach ended up being... but it was new.

And without bent/ folded over edges they need a straight eased (not cut you sharp) perimeter on the entire piece. Unless covered with plastic moldings like the wheel wells.

Yeah, look at it whiz through that part. 3 1/2 minutes with time lapse. Hope Tesla has more than one brake press.

It's a demo at a trade show. How many pallets of parts did you want them to make?

 

[Singuy]

Do some math. $30K for an H100, which Elon says is the equivalent of 3 A100s. Tesla published how many A100-equivalents they will have online next year. This is straightforward.

I'll be extremely disappointed if Tesla keep buying billions of dollars of H100 especially their diminishing in return hits hard with large NN training and D1 is suppose to be cheaper and fix most of Nvidia's short comings.

 

[JupiterMan]

I'll be extremely disappointed if Tesla keep buying billions of dollars of H100 especially their diminishing in return hits hard with large NN training and D1 is suppose to be cheaper and fix most of Nvidia's short comings.

If buying more compute short term gives Tesla the compute to progress FSD quicker and also give us (at last!) international releases, then the benefits to Tesla and to us shareholders are huge.

Dojo has to scale, and I don't think Tesla should sit on their hands waiting for that.

Also gives capacity for Optimus which is probably not getting much compute set aside for it currently.

I'm happy with this.

Disclosure: I also hold NVDA.

 

[Mengy]

Actually, large pieces of stainless are bent routinely. Ever look at most elevator doors? Bent perfectly with no ripple, etc. probably Elon got the idea for the cybertruck in an elevator.

Nope. I worked as a sheetmetal engineer for decades and routinely used SS, and no elevator doors are "perfect". You don't need high precision tolerances on something like a flat elevator door, and if you actually checked the tolerances and finish on those doors you'd find lots of variations because it's not a precision product. Not anything like what Elon is calling for on the CT.

Tesla will need to fight with lots of variables to form the CT panels with that kind of precision. Springback, variations in metal thickness, grain direction, etc. Elevator doors would be a cakewalk compared to what the CT will require in those kinds of volumes.

 

[Singuy]

If buying more compute short term gives Tesla the compute to progress FSD quicker and also give us (at last!) international releases, then the benefits to Tesla and to us shareholders are huge.

Dojo has to scale, and I don't think Tesla should sit on their hands waiting for that.

Also gives capacity for Optimus which is probably not getting much compute set aside for it currently.

I'm happy with this.

Disclosure: I also hold NVDA.

I ain't complaining about their current h100 purchase because short term is fine. 2 years from now is not fine and it's no longer short term. 2 years in computer years is a long ass time especially Dojo is in full production. Tesla spent hundreds of millions on Dojo to save a few billion from buying Nvidia..at least that was how it was sold to investors.

 

[hitchhiker]

There's no such thing, that's was made up for the film adaptation of LoTR, along with other awful ideas like "dwarf tossing"

Visually impressive films, but they destroyed the source material... and don't get me started on The Hobbit films

Elen síla lúmenn’ omentielvo...

Tell us how you feel about the Rankin/Bass animated musical versions from the late 70s?

 

[AudubonB]

Nipping the LoTR posts right now before they get out of hand. They do not belong here.

 

[SOULPEDL]

I wonder if ARK started buying back some TSLA?
They saw it coming last time...

Spoiler: ARKK sold 4 times Aug 10-15.

 

[Buckminster]

Threads of the day:
4680 cell design, chassis integration & factory discussion for investors Jordan on M3P
FSD V12 (end to end AI) V12 in training
Project Dojo - the SaaS Product?

https://twitter.com/x/status/1696914526313107513

Range looks good until you see the pack size.

 

[Max Plaid]

Interesting, this video of a Model 3 Hghland shows what appears to be the trunk opening much higher up than the current car, could this actually be a hatch now?

https://twitter.com/x/status/1696911108068773998

 

[JupiterMan]

I ain't complaining about their current h100 purchase because short term is fine. 2 years from now is not fine and it's no longer short term. 2 years in computer years is a long ass time especially Dojo is in full production. Tesla spent hundreds of millions on Dojo to save a few billion from buying Nvidia..at least that was how it was sold to investors.

Elon has a good history of not allowing sunken cost fallacy in his businesses.

He's probably hedging bets on DOJO right now, and if it doesn't deliver (don't get me wrong I think it will) then don't be surprised if they dump it / start over again.

This is part of what makes Elon Elon.

Best to think about the future returns on FSD...... Nvidia chip cost is inconsequential.

 

[JupiterMan]

Interesting, this video of a Model 3 Hghland shows what appears to be the trunk opening much higher up than the current car, could this actually be a hatch now?

https://twitter.com/x/status/1696911108068773998

looks the same place to me?

 

[Krugerrand]

It's a demo at a trade show. How many pallets of parts did you want them to make?

Don’t know. Don’t care.

The point I was making is that there is in fact a limit to speed of bending. The big presses Tesla uses for steel and aluminum body panels top out at like 20/22 strokes per minute and Tesla is not running the presses at that speed. Far more likely at 12-15 strokes per minute.

How fast can a bending machine bend the various Cyber panels considering their size, weight, type of bend etc….?

Like I said, hope they have more than one brake press or that there’s an extreme simplicity to the bends they need and that the final composition of the material adds to that simplicity. I suspect, and given Elon’s comments, it isn’t so.

 

[Krugerrand]

Interesting, this video of a Model 3 Hghland shows what appears to be the trunk opening much higher up than the current car, could this actually be a hatch now?

https://twitter.com/x/status/1696911108068773998

Let me close one eye and squint with the other… Oh, yeah. I can see it now. Totally in a higher spot - not.

 

[Artful Dodger]

I'll be extremely disappointed if Tesla keep buying billions of dollars of H100 especially their diminishing in return hits hard with large NN training and D1 is suppose to be cheaper and fix most of Nvidia's short comings.

If they were smart, they'd fund that H100 cluster with NVDA calls bought BEFORE announcing $BILLIONS spent on H100s...

Cheers!

 

[StarFoxisDown!]

I usually don't bother to read into open interest/volume charts for options for the following week and the week after that since most of the time the open interest and volume's are so low in numbers that they will easily change greatly at the start of the next week.

But as I was looking at options for this week to see how much incentive there is for MM's to cap at 260 or try and drop the stock to 249.99 by Friday, I happened to look at next week's options and the week after and there's some noticeable updates/differences between when I looked a couple days ago and today. Namely, that there's a big call wall forming for 300 next week with decent amount of volume forming for Calls between 250-300.

But it's the following week that's really interesting. A big call wall at 400/share. Granted, some of these calls are LEAPS. So bets made a long time ago that are still open. But the noticeable thing is that just a couple days ago, this call wall at 400/share was much, much smaller. I can't remember the exact number it was at but I think it's at least doubled in size in the past few days.


I don't want to read too much into that, but clearly there's some big bets coming in that TSLA will be much higher, well clear of 300 in two weeks.

And obviously, don't take this as go out and YOLO TSLA. This could be strategic hedging or could Wall St just playing games to make it looks like there's bullish bets on the options only to pull the rug. We've seen it happen before.

 

[Artful Dodger]

CATL M3P // Is the Hype Legit & Will Tesla Use it? | The Limiting Factor (1 hr ago)

 

[Buckminster]

Must read post from Jordan here:
4680 Cathode & Anode discussion for investors
Potential Jeff Dahn breakthroughs in Sodium, Silicon Carbide, Tesla LFP chemistry and NMC

 

[SOULPEDL]

Let me close one eye and squint with the other… Oh, yeah. I can see it now. Totally in a higher spot - not.

I can see the trunk line and their black cover is tucked into it. Same location as M3.

Highland / Model 3 Shown.

 

[UkNorthampton]

For comparison - June 23 UK tesla Y sales were around 5500 (from China, high month, end of quarter).

5000 Tesla Y to Turkey from mid-May to now, 8000 expected by end of September.

No way to investigate claim, these numbers aren't being counted in estimates:-
Translated

"Will Tesla break the world record in Turkey

I learned that Tesla has delivered around 5,000 Model Ys in Turkey so far.

It was expected that the number of deliveries would reach 8,000 by the end of September, with more than 3,000 vehicles towed to Istanbul Park a short while ago.

In October, the target was to exceed 10,000 deliveries.

Delivery of 10,000 units of a model every 5 months could be a world record for Tesla..."

https://twitter.com/x/status/1696929415169528056