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

[Cherry Wine]

I enjoy following TSLAQ because many of its members are professionals that have access to data that retail does not. Occasionally, they will post some of this data but interpret it incorrectly because they are not tuned into what Tesla is working on as the bulls are. Ironically, they often boost my confidence in my investment.

I can no longer find the Tweet (probably deleted because he probably realized it was bullish later on), but today TSLAQ member Keubiko posted a chart about battery shipments from Kinki region of Japan to the US. It showed that in September the shipments to the US dropped to zero, while shipments to China continued at the same pace. He asked the question “Is Tesla discontinuing S/X?”

I think this potentially means all of S/X are moving to the new roadrunner cells (or 2170, which I find unlikely). It also means production has likely already begun and Plaid S/X are imminent. I’ve heard some rumblings from the moderator of RealTesla before that Tesla was going to increase S/X production in the past. We could potentially see a large price drop in the price of the S/X while the Plaid will be the super high end product.

It was this tweet. Roadrunner for S/X was my first thought as well. It's still live here, by the way:

https://twitter.com/KawasakiKR11/status/1306125252636880897?s=20

 

[dl003]

a week ago I reported that TSLA might have switched to a new motherboard for S and X. Alien technology anyone?

 

[Cherry Wine]

a week ago I reported that TSLA might have switched to a new motherboard for S and X. Alien technology anyone?

Right - a supplier reported a big drop in orders, or something?

 

[Nurkbonis]

I saw that tweet. Excellent analysis. It makes a lot of sense.

I enjoy following TSLAQ because many of its members are professionals that have access to data that retail does not. Occasionally, they will post some of this data but interpret it incorrectly because they are not tuned into what Tesla is working on as the bulls are. Ironically, they often boost my confidence in my investment.

I can no longer find the Tweet (probably deleted because he probably realized it was bullish later on), but today TSLAQ member Keubiko posted a chart about battery shipments from Kinki region of Japan to the US. It showed that in September the shipments to the US dropped to zero, while shipments to China continued at the same pace. He asked the question “Is Tesla discontinuing S/X?”

I think this potentially means all of S/X are moving to the new roadrunner cells (or 2170, which I find unlikely). It also means production has likely already begun and Plaid S/X are imminent. I’ve heard some rumblings from the moderator of RealTesla before that Tesla was going to increase S/X production in the past. We could potentially see a large price drop in the price of the S/X while the Plaid will be the super high end product.

 

[mrdoubleb]

I enjoy following TSLAQ because many of its members are professionals that have access to data that retail does not. Occasionally, they will post some of this data but interpret it incorrectly because they are not tuned into what Tesla is working on as the bulls are. Ironically, they often boost my confidence in my investment.

I can no longer find the Tweet (probably deleted because he probably realized it was bullish later on), but today TSLAQ member Keubiko posted a chart about battery shipments from Kinki region of Japan to the US. It showed that in September the shipments to the US dropped to zero, while shipments to China continued at the same pace. He asked the question “Is Tesla discontinuing S/X?”

I think this potentially means all of S/X are moving to the new roadrunner cells (or 2170, which I find unlikely). It also means production has likely already begun and Plaid S/X are imminent. I’ve heard some rumblings from the moderator of RealTesla before that Tesla was going to increase S/X production in the past. We could potentially see a large price drop in the price of the S/X while the Plaid will be the super high end product.

Agreed, but before we get ahead of ourselves, it could just be that they only have data on full months and September was zero because it has not ended yet. The Tweet (which is quoted above by @Cherry Wine) does show a big drop in August though and, as you say, it does suggest the opposite of what they were reading into it. The only question is will all S/X get the new cells or just Plaid and whether they will get Roadrunner or the new, improved 2170s they have recently started producing on the former TE cell line at Giga Nevada.

 

[gluu]

It was this tweet. Roadrunner for S/X was my first thought as well. It's still live here, by the way:

https://twitter.com/KawasakiKR11/status/1306125252636880897?s=20

Thank You, I had the two users confused.

 

[kengchang]

Agreed, but before we get ahead of ourselves, it could just be that they only have data on full months and September was zero because it has not ended yet. The Tweet (which is quoted above by @Cherry Wine) does show a big drop in August though and, as you say, it does suggest the opposite of what they were reading into it. The only question is will all S/X get the new cells or just Plaid and whether they will get Roadrunner or the new, improved 2170s they have recently started producing on the former TE cell line at Giga Nevada.

It doesn't make whole lot of sense for S/X to have two battery pack configuration of RoadRunner and 2170. Giga Nevada has been using imported cell for Powerwall/MegaPack so there's no spare capacity of 2170 to go into S/X

 

[Bet TSLA]

For weeks on end this entire year, people here have accurately predicted the closing share price on Fridays........for weeks.

Options trading and MM's pushing the stock one way or the other for their benefit is not conspiracy.

Prove it. The history is available. I suppose it could just be people I have on ignore due to my perception that they generate little other than noise. Or maybe there are a bunch of different predictions and you just remember the ones that are right. People do predict all sorts of crap, then forget what they got wrong and remember what they got right. And, as has been pointed out many times, anybody who can actually predict prices at specific times can become rich really fast. And yet... nobody does.

Go ahead. Tomorrow you can predict the closing price at least a couple of hours before, preferably when it isn't actually near that price. Then load up on the best options you can find. Should be trivial to multiply your money by four or five. Post about it. Be a hero! I'll happily predict failure. Actually, I predict you won't do it because you don't actually believe.

 

[gluu]

a week ago I reported that TSLA might have switched to a new motherboard for S and X. Alien technology anyone?

Thanks for this. Unclear which motherboard they currently provide, but if its the MCU board...

I've been speculating that Tesla will combine the MCU and Autopilot boards to reduce costs. They already have to talk to each other to perform sentry mode. If the motherboard is also getting upgraded soon then this report makes more sense.

Tesla is working on HW 4.0 self-driving chip with TSMC for mass production in Q4 2021, report says - Electrek

From the article:

Each 12-inch wafer can only be cut out. 25 chips. Production of the new chips will begin in the fourth quarter, with an initial production of about 2,000 wafers, and it is expected to enter full mass production after the fourth quarter of next year.”

2000 wafers * 25 chips in Q4 = 50,000 SoW (Systems on Wafer). I'm uncertain on what yields we are likely to see, but I imagine they are quite high for this process since the substrate for the chips is a much older process node (~14nm I believe). If we assume even 50% yield, we are looking at S/X going back to full production of 90-100k units/year.

At Autonomy Day, 4/2019, Elon said HW4 was 1 year into a 3 year development cycle. We could potentially be seeing HW4 in production for S/X soon although it seems a bit early still.

Don't want to get my hopes up, but we could be seeing a really advanced vehicle very soon, Battery Day Tech + HW4 + S/X redesign.

 

[StarFoxisDown!]

Prove it. The history is available. I suppose it could just be people I have on ignore due to my perception that they generate little other than noise. Or maybe there are a bunch of different predictions and you just remember the ones that are right. People do predict all sorts of crap, then forget what they got wrong and remember what they got right. And, as has been pointed out many times, anybody who can actually predict prices at specific times can become rich really fast. And yet... nobody does.

Go ahead. Tomorrow you can predict the closing price at least a couple of hours before, preferably when it isn't actually near that price. Then load up on the best options you can find. Should be trivial to multiply your money by four or five. Post about it. Be a hero! I'll happily predict failure. Actually, I predict you won't do it because you don't actually believe.

Seriously, do your own leg work. I'm not here to convince you.

Dear lord . Feel free to disagree with me but try not to be an *ss

 

[mrdoubleb]

It doesn't make whole lot of sense for S/X to have two battery pack configuration of RoadRunner and 2170. Giga Nevada has been using imported cell for Powerwall/MegaPack so there's no spare capacity of 2170 to go into S/X

We kind of agree on the first part, it probably just depends on the volume they can produce right now. However if the new pack can't incorporate two different types of cells, there is no way they would have 2 different packs as it may also require chassis modifications.

As for Nevada cells, Carsonight, who is usually a reliable source on these things, reported a few days ago, that they have converted the one and only TE line at GF1 for vehicle cells and that one line is now producing brand new, higher density and longer lasting ones. This seems to corroborate other reports of Tesla introducing not one, but at least 2 new cells/chemistries next week - one would be a next gen 2170 (an evolution) the other Roadrunner (revolution). My thinking is, one of these products, the higher volume one, goes into Palladium/Plaid S/X, while the other will first show up in the Semi and/or the Roadster.

 

[mrdoubleb]

Thanks for this. Unclear which motherboard they currently provide, but if its the MCU board...

I've been speculating that Tesla will combine the MCU and Autopilot boards to reduce costs. They already have to talk to each other to perform sentry mode. If the motherboard is also getting upgraded soon then this report makes more sense.

Tesla is working on HW 4.0 self-driving chip with TSMC for mass production in Q4 2021, report says - Electrek

From the article:



2000 wafers * 25 chips in Q4 = 50,000 SoW (Systems on Wafer). I'm uncertain on what yields we are likely to see, but I imagine they are quite high for this process since the substrate for the chips is a much older process node (~14nm I believe). If we assume even 50% yield, we are looking at S/X going back to full production of 90-100k units/year.

At Autonomy Day, 4/2019, Elon said HW4 was 1 year into a 3 year development cycle. We could potentially be seeing HW4 in production for S/X soon although it seems a bit early still.

Don't want to get my hopes up, but we could be seeing a really advanced vehicle very soon, Battery Day Tech + HW4 + S/X redesign.

Doesn't that article say it is 7nm though? Power requirements and heat dissipation was an important factor for HW3, they have talked about it during Autonomy Day, so if they go up in processing power with HW4, one would think they would want to go 7nm to keep power requirements under control.

 

[gluu]

We kind of agree on the first part, it probably just depends on the volume they can produce right now. However if the new pack can't incorporate two different types of cells, there is no way they would have 2 different packs as it may also require chassis modifications.

As for Nevada cells, Carsonight, who is usually a reliable source on these things, reported a few days ago, that they have converted the one and only TE line at GF1 for vehicle cells and that one line is now producing brand new, higher density and longer lasting ones. This seems to corroborate other reports of Tesla introducing not one, but at least 2 new cells/chemistries next week - one would be a next gen 2170 (an evolution) the other Roadrunner (revolution). My thinking is, one of these products, the higher volume one, goes into Palladium/Plaid S/X, while the other will first show up in the Semi and/or the Roadster.

To further tie in one more element, this article mentions another GigaPress, which everyone assumes is for Model Y:

Tesla is deploying a second Giga Press machine in (or actually out of) Fremont factory - Electrek

Could it be that this is for S/X instead? Are two GigaPresses needed for Tesla to reach their target Model Y volumes out of fremont?

 

[gluu]

Doesn't that article say it is 7nm though? Power requirements and heat dissipation was an important factor for HW3, they have talked about it during Autonomy Day, so if they go up in processing power with HW4, one would think they would want to go 7nm to keep power requirements under control.

My research into this topic was that the process, TSMC's InFo_SoW is a method of using a wafer as a motherboard if you will. The 7nm chips would be fabbed as usual but then somehow attached to a 14nm wafer that has the appropriate circuit layout. The 14nm wafer replaces the motherboard, not the 7nm HW4 Chip and related SoC components.

EDIT: TSMC to Build Supercomputing AI Chips, Ramps Wafer-Scale Computing - ExtremeTech

The above link should help clarify, please let me know if you interpret it differently.

 

[JBRR]

Tesla Wins Case Against Former Employee Accused Of Hacking, Transferring Data

 

[Todesbuckler]

China August EV sales by ev-sales blog are online.

#1 – Tesla Model 3

Looking at the manufacturers ranking, BYD (15%) is still in the leadership, but the #2 Tesla (13%), and especially the rising #3 SAIC (13%, up 1%) are looking to catch it soon.

Link: EV Sales: China August 2020

 

[Discoducky]

Nice to see you posting....IIRC you have not been on 'investors' for awhile?

Yeah, it has been awhile, thanks for noticing! I absolutely love how many amazing folks post on this thread; you included .

The insights and discussions are invaluable, but my investment accounts have grown so much maybe I can actually value them

Had to share some thoughts on battery day as it maybe the biggest tech announcement since...Model S

 

[MC3OZ]

I've been speculating that Tesla will combine the MCU and Autopilot boards to reduce costs.

Yes, that is also how I read HW4 it seems like a "System on Chip" rolling a lot of stuff into a single chip.

 

[Tslynk67]

fwiw I made a max pain script that runs every hour. Will be interesting to see how it evolves tomorrow during the trading hours. The output can be seen here: http://home.pusen.org/stocks/maxpain

Nice, but, erm...

Max pain for expiry 2020-09-18 is $999
Max pain for expiry 2020-09-25 is $999
Max pain for expiry 2020-10-02 is $999
Max pain for expiry 2020-10-02 is $999
Max pain for expiry 2020-10-16 is $999
Max pain for expiry 2020-10-23 is $1
Max pain for expiry 2020-10-30 is $1
Max pain for expiry 2020-11-20 is $999
Max pain for expiry 2020-12-18 is $999
Max pain for expiry 2021-01-15 is $999
Max pain for expiry 2021-02-19 is $999
Max pain for expiry 2021-03-19 is $915
Max pain for expiry 2021-06-18 is $900
Max pain for expiry 2021-07-16 is $999
Max pain for expiry 2021-09-17 is $999
Max pain for expiry 2022-01-21 is $750
Max pain for expiry 2022-03-18 is $999
Max pain for expiry 2022-06-17 is $810
Max pain for expiry 2022-09-16 is $999
updated: 2020-09-18 08:00:18.825098+02:00

 

[chiller]

Keep in mind that Max pain is dependent on open interest. While volume is getting updated in real time the OI only gets updated once every day.

While the script is nice I don’t know how much information you can get from the hourly updates. Plus based on my experience max pain usually matters for monthly options not so much for the weekly options.

Thanks for that info.
Thats what I get for coding when the market is closed.
Guess the hourly updates will be meaningless.