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

[nativewolf]

@mongo If they plan to put a 2170 line in Austin the 4680 ramp is going badly. By Q3 they were planning to be running Austin on Austin production. Not happening. CT with no 4680? Nope. Semi with no 4680? nope. Retrofitting old battery cell formats into Austin speaks to Tesla being fairly agile but it doesn't speak well to the new cell production processes.

I need a CT. Not I want a CT. I need it. Actually our company needs 3 of them. Sucks for us. We had a discussion last week about alternatives because we don't think the current vehicles are going to make it another year. Sucks...sucks sucks. We may have to buy 3 ICEs. Sucks.

 

[Usain]

What are you smoking? They built a factor to explicitly use a cell size that they can't make in volume. I mean it is one thing to be a fanboy but another to have your head in the ground. They are not happy. They are not seeing success coming so they will retrofit 2170 into Austin. That's a huge fail, and that sucks for the EV transition. They wont produce CT with 2170 nor the Semi. This is an issue.

Time to chill, folks. This is Dylan Loomis who got it from someone inside Tesla who remains anonymous. We don't really know how 4680 production is going. You could already infer from the last earnings call that Tesla would like to be further along, and that's all Dylan's source was saying. So it's not really news.

Actually, I found Dylan's video to be mostly positive. One million 4680 cells per month is a good jump from the relatively recent report from Kato Road that Tesla had produced only one million 4680's total. And at least the cells are going into vehicles and being sold to the public.

This is not what a failed technology looks like. This is what a ramping-up technology looks like.

It's still the first inning for 4680. We don't need a grand slam in the first to win the game.

 

[TheTalkingMule]

This is why YouTuber stuff should be kept to dedicated threads. If some jackhole wants to spout "inside information", it'd be nice if it didn't derail this thread.

 

[damonbrodie]

An issue that doesn't exist besides some rehash rumors from "sources". None of these concerns were expressed during the conference call. "Cybertruck is on track for production next year" said by Elon. "Gigatexas is spooling up" said by Elon via email today. 2170 line going into Gigatexas from that same earnings call because they have too many batteries and they want to make LR AWDs as it's very popular.

But hey if you think Texas is only capable of making 40 cars a day, and constantly have 7 days worth of production sitting in their logistic lot from every fly over then sure....

This is why YouTuber stuff should be kept to dedicated threads. If some jackhole wants to spout "inside information", it'd be nice if it didn't derail this thread.

By all means lets go back to talking about bouncing cats and shoulders and heads then?

I welcome all news both good and bad. I have a lot of TSLA (for my meager means) riding on it so I want the most up to date info. I could use more information like that than the size of the candles at any particular moment.

 

[mongo]

Listen - all I'm saying is that it isn't ramping as quickly as some of us might have been expecting. 1M cells per months is great, but it needs to be an order of magnitude larger or more before its meaningful. I'm not saying their progress isn't great, but from an investor perspective there is clearly risk on the way to production scaling and that scale is not necessarily a sure thing. This forum should be more than just a fan club. I'd hope we can have intelligent discussions on that makes the stock go both up and down.

I still stand by my comment - it looks like the 4680 is 1y away from full production. Please provide evidence to the contrary if you disagree.

Worst case, If they have X lines that make 1 million cells a month, then all they need to make 10 million cells a month is 10 times as many lines. That's a problem money and lead time solves, as opposed to a physical impossibility.

Good question - how many MY SR do we expect Tesla to ship per month? 5000, 8000, 10000, more? Somewhere in that ball park. Right now they are at 1200-1500 per month.

Edit: Ideally they would get 4680s cranking out enough to switch the MYP and MY LR over to 4680s then they would enough cells for 50K cars per month at Austin.

Just like S/X with 18650, Fremont is likely not going to switch to 4680 as long as 2170 supply exists.

@mongo If they plan to put a 2170 line in Austin the 4680 ramp is going badly. By Q3 they were planning to be running Austin on Austin production. Not happening. CT with no 4680? Nope. Semi with no 4680? nope. Retrofitting old battery cell formats into Austin speaks to Tesla being fairly agile but it doesn't speak well to the new cell production processes.

I need a CT. Not I want a CT. I need it. Actually our company needs 3 of them. Sucks for us. We had a discussion last week about alternatives because we don't think the current vehicles are going to make it another year. Sucks...sucks sucks. We may have to buy 3 ICEs. Sucks.

Disagree.
Capability is not the same as requirement. Berlin already paved the way equipment wise, it's about flexibility and risk mitigation.

For perceptive, the first Model Y Austin deliveries were in Q2 (2 months ago). From the Q1 transcript I quoted, they expect 4680 in volume in Q4, maybe late Q3.

 

[2daMoon]

Because Pano is realistic. The entire falling out a couple of years ago was, I suspect, business related and based on very different cultures. Tesla wanted to go 4680 fast and Pano probably said not possible. Might also have been costs..or a combination. It's clear that they thought they'd have enough 4680 to have Austin running on just 4680 and now that isn't going to happen. So they will retro fit 2170. It is so very very very clear to anyone looking that it is not going well. It's the most important thing in the entire EV industry (battery production) and that's what (IMO) Tesla should have been laser focused on. It's the single biggest gating factor to the EV transition and solar energy as well.

It also means that OEMs move a couple of years closer to Tesla in terms of functional battery capacity. Last year I kept looking for battery plant announcements that never came..maybe this is why.

It isn't breaking news today if they told us they were going to do this at the last earnings call.

Anyone who followed what was presented is neither surprised, nor disappointed. They have been transparent.

Maybe you haven't heard, though it seems Elon has hinted at least one time, if not more...

PRODUCTION IS HARD
​

 

[TheTalkingMule]

Because Pano is realistic. The entire falling out a couple of years ago was, I suspect, business related and based on very different cultures. Tesla wanted to go 4680 fast and Pano probably said not possible. Might also have been costs..or a combination. It's clear that they thought they'd have enough 4680 to have Austin running on just 4680 and now that isn't going to happen. So they will retro fit 2170. It is so very very very clear to anyone looking that it is not going well. It's the most important thing in the entire EV industry (battery production) and that's what (IMO) Tesla should have been laser focused on. It's the single biggest gating factor to the EV transition and solar energy as well.

It also means that OEMs move a couple of years closer to Tesla in terms of functional battery capacity. Last year I kept looking for battery plant announcements that never came..maybe this is why.

That falling out was purely about money. Panasonic wanted a deeper cut and was willing to bottleneck cell production for some sort of negotiation leverage.

Didn't work out so hot.

 

[damonbrodie]

Worst case, If they have X lines that make 1 million cells a month, then all they need to make 10 million cells a month is 10 times as many lines. That's a problem money and lead time solves, as opposed to a physical impossibility.

True. I suppose they have contingency plans in place if they can't get the scale per line to their liking.

Just like S/X with 18650, Fremont is likely not going to switch to 4680 as long as 2170 supply exists.

I understand the plan is to move the MY (all variants) to 4680s once they get the cell scale up. All MYs would ship out of Texas and thus avoid any Osbourne effects on the MY.

Disagree.
Capability is not the same as requirement. Berlin already paved the way equipment wise, it's about flexibility and risk mitigation.

For perceptive, the first Model Y Austin deliveries were in Q2 (2 months ago). From the Q1 transcript I quoted, they expect 4680 in volume in Q4, maybe late Q3.

Agreed - that's the plan they published. Is it still on track though?

 

[damonbrodie]

That falling out was purely about money. Panasonic wanted a deeper cut and was willing to bottleneck cell production for some sort of negotiation leverage.

Didn't work out so hot.

The fact that Tesla shared the 4680 formula with Panasonic tells me it isn't purely about money. I think Panasonic wasn't innovating new cell tech fast enough.

 

[TheTalkingMule]

I heard these guys don't even test the brakes before delivering EVs. Do we have any proof they're doing so at the new Austin plant?

You never know!

 

[2daMoon]

The fact that Tesla shared the 4680 formula with Panasonic tells me it isn't purely about money. I think Panasonic wasn't innovating new cell tech fast enough.

I don't think that Panasonic is using the Dry Electrode process. I'd expect that they designed their production line to use the same form factor and use the tabless technology.

There are several discrete innovations that Tesla is refining for the 4680 which aren't necessarily being developed or implemented by Panasonic or others, because they will build batteries for Tesla based upon what they are familiar with and equipped for. Going from wet to dry electrode would be a hurdle for them. Not that it isn't for Tesla too, they are just closer to accomplishing it at scale.

 

[mongo]

I understand the plan is to move the MY (all variants) to 4680s once they get the cell scale up. All MYs would ship out of Texas and thus avoid any Osbourne effects on the MY.

What? Where is that coming from?

That doesn't make sense to shutter the FremontbY lines.

Agreed - that's the plan they published. Is it still on track though?

It's not even Q3 yet, the call was 1.5 months ago. At only 25%-33% through the time span, why believe that it isn't?

The fact that Tesla shared the 4680 formula with Panasonic tells me it isn't purely about money. I think Panasonic wasn't innovating new cell tech fast enough.

What formula? Do we know Panasonic is not doing wet electrode?

 

[⚡️ELECTROMAN⚡️]

What are you smoking? They built a factor to explicitly use a cell size that they can't make in volume. I mean it is one thing to be a fanboy but another to have your head in the ground. They are not happy. They are not seeing success coming so they will retrofit 2170 into Austin. That's a huge fail, and that sucks for the EV transition. They wont produce CT with 2170 nor the Semi. This is an issue.

Are they making the 2170's at Austin or just changing the production line to build cars with 2170's, that are manufactured elsewhere?

 

[damonbrodie]

What? Where is that coming from?

That doesn't make sense to shutter the FremontbY lines.

Not shutter - use them to increase M3 capacity. I remember this from one of Elon's videos. I could be wrong - I'll see if I can find it.

It's not even Q3 yet, the call was 1.5 months ago. At only 25%-33% through the time span, why believe that it isn't?


What formula? Do we know Panasonic is not doing wet electrode?

No idea on the dry/wet. I am probably mistaken - I thought all 4680s would be the same.

 

[Singuy]

True. I suppose they have contingency plans in place if they can't get the scale per line to their liking.



I understand the plan is to move the MY (all variants) to 4680s once they get the cell scale up. All MYs would ship out of Texas and thus avoid any Osbourne effects on the MY.


Agreed - that's the plan they published. Is it still on track though?

No one said the plan is to move all variants to 4680s. Maybe in the distant future when there are way more 4680 lines than any other line but as of today, if you want a LR Model Y, it has to be on the 2170s. So if they want to add a 2170 line, then they just want to build more LR Model Ys. Understand that SR Model Y which will be a few thousand cheaper will be the most popular model world wide like the SR model 3 is. But because the SR Model Y has just begun to ramp, the majority of Model Y orders will still be the LR(especially when you can't even order a SR Model Y). When you have a year worth of back orders of the LR, and you don't want people to cancel their orders for the SR Model Y, it's best to keep this invitation only and add a 2170 line where ever there's space.

 

[3'sterly]

Are they making the 2170's at Austin or just changing the production line to build cars with 2170's, that are manufactured elsewhere?

Exactly. I think the latter. Makes the most sense. Build it and they will sell.
Optimize the 4680 battery on it’s own timeline.

 

[Singuy]

I don't know why there's so much confusion. When you have a year worth of backlog for the Model Y LR and there's no supply shortages of 2170s, every new 2170 line is a money printer. I'm pissed that Tesla didn't put in a 2170 first at Texas.

 

[MP3Mike]

They are cell constrained at 1200cells per month (EDIT that should be cells for 1200 MR SR vehicles per month. - roughly 1M cells per month.) for the MR SR.

I was referring to what Tesla said on the last earnings call. As a whole they are not cell constrained. They can only make x total vehicles across all factories. For every 4680 Model Y they make at GigaTexas that is one less vehicle they can make a different factory. (Limited by chip and other parts.) For example in April/May they were shipping cars without charge port ECUs for a few weeks.

I think that the LR Model Y has better margins, so they want to maximize how many of them they make, and use all of the "left over" parts for making Model Y AWDs at GigaTexas.

Thus they aren't cell constrained. i.e. if they had more cells they wouldn't make more vehicles in total. (Sure maybe they could make more at GigaTexas than they are currently, but it is probably better to take the ramp nice and slow to work out all the kinks and refine things before you go full speed.)

 

[MP3Mike]

I don't know why there's so much confusion. When you have a year worth of backlog for the Model Y LR and there's no supply shortages of 2170s, every new 2170 line is a money printer. I'm pissed that Tesla didn't put in a 2170 first at Texas.

Why are you pissed? From what they said there are plenty of 2170s, but not enough of everything else to make more cars. So putting a line in at GigaTexas to make cars with 2170 packs, sourced from GigaNevada, wouldn't result in more total cars being built. But it would delay the 4680 structural pack/Model Y ramp. Using the time that other parts are limiting total production is the perfect time to start a new production process. (The more experience they get making them, and the more customer driven miles driven in the new cars, the better.) When the chip, and other part, supply issues are resolved they will be able to ramp the 4680 line more. (And they will probably have 2170 capability at GigaTexas before those part shortages resolve.)

 

[2daMoon]

Why are you pissed? From what they said there are plenty of 2170s, but not enough of everything else to make more cars. So putting a 2170 line in at GigaTexas wouldn't result in more total cars being built. But it would delay the 4680 structural pack/Model Y ramp. Using the time that other parts are limiting total production is the perfect time to start a new production process. (The more experience they get making them, and the more customer driven miles driven in the new cars, the better.) When the chip, and other part, supply issues are resolved they will be able to ramp the 4680 line more. (And they will probably have 2170 capability at GigaTexas before those part shortages resolve.)

The way I understood it, they would use 2170 packs in cars built at Giga Texas (and Germany) if necessary while ramping 4680 production, not establish any new production line for 2170 batteries themselves. (maybe that is what you are saying here)