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

[Bladerskb]

Great, then please provide us the right information.

Sure.

"Currently even though Tesla haven't solved FSD, Legacy auto makers have a hard time buying from Nvidia or other companies in scale due to the prohibited cost. Nvidia's L5 TRAINING platform is almost 10k, and this doesn't include the 8 or so lidars needed. Not to mention an entire cooling solution that need to dissipate about 3kw of power. Even so FSD isn't solved using this 30k-40k hardware package."​

Other companies chip doesn't use 3kw, you are using figures from 10 years ago. All self driving companies either use Nvidia, Mobileye, Qualcomm, Huawei or developed their own. Lets take for example NIO ET7 releasing Q1 2022. It uses 4x Nvidia Orin which provides 1,096 TOPs of power while using only ~180 watts. Each Orin chip is 258 Tops/45 watts.

This is compared to HW3's 144 Tops/ 100 watts, almost 8x more powerful with only 1.8 more wattage.
That's way more powerful and efficient than FSD Computer while also being smaller.

Now about the Lidars, its also a myth that Lidars are power hungry.

Livox 12 Watts
https://livoxtech.com/horizon/specs
Luminar 15 Watts
https://wired.com/story/lidar-cheap-make-self-driving-reality/…
Innovision under 40 Watts
https://oemoffhighway.com/electronics/sensors/product/21073236/innovusion-inc-innovusion-cheetah-lidar-system…
Innoviz One 15-23 Watts
https://just-auto.com/interview/ces-israeli-start-up-bags-capital-accelerates-lidar-innovation_id192792.aspx…

All listed lidar are high resolution & will go into mass production on cars this year & next year.
What about lidar costs? All the lidar listed above cost well under $500.

"Currently even though Tesla haven't solved FSD, Legacy auto makers have a hard time buying from Nvidia or other companies in scale due to the prohibited cost."​

To go even more into details not only will NIO deploy Nvidia, Volvo will also deploy 1x Orin in 2022 in their Volvo XC90.
2022 BMW iX will use 2x of Mobileye's next gen EyeQ5 (48 Tops/20 watts).
Geely's 2021 Zeekr 001 will use Mobileye's complete supervision (2x EyeQ5).
BAIC's 2021 Arcfox will use Huawei's ADS with their MDC 810 computer (400/800 TOPS).

I could go on and on.

​

"No auto company will copy this as they wouldn't even know where to start. Nvidia, Mobile eye, and Waymo might ditch their approach asap trying to catch up, which is a totally different strategy as they mainly focused on localization using Lidar and HD maps."​

The myth that Lidar and/or HD maps, slows you down or prevents you from scale needs to die.

There are two system's releasing by the end of 2021 that will have door to door city streets autonomy anywhere in china.
Remember that china is orders of magnitude harder to drive than in the US.

2021 Zeekr's 001 with Mobileye's Supervision
2021 Arcfox Alpha S with Huawei's ADS

 

[petit_bateau]

I dare you to get a project dropping electric cables along populated coasts in Europe to be a quick project. Or major new ones over land for that matter.

Teslas issues with environmental permits for the Berlin factory wouldn't even register on that scale.

Actually. Wonder what the boring company would charge for a much smaller tunnel than they are doing now. Probably the cheapest. Almost certainly the fastest in many areas. Except there would be protests about damaging the groundwater etc.

Sigh, wonder how many solarpanels I can fit on my apartment balcony. I might have to soon.

In general terms running the major grid interconnects subsea is exactly what Europe is doing. It doesn't matter too much whether they are straight HV(AC) or HVDC though increasingly the latter is winning. The grid-shore connections are somewhat problematic but ordinarily solvable, generally including short sections of underground through the most NIMBY bits.

Running overland for long distances in Europe is very problematic, and going underground is uneconomic. That is why only short underground sections are used. It is not just the cost of making the hole that drives up the price, there are other factors to do with the kit. This is why the N-S trans-German links to connect Baltic wind with Bavaria are so contentious.

In China, Brazil, India, etc overland is doable, but even they go underground for short sections going into major city centres unless access corridors were left unbuilt, and ordinarily they weren't. Real estate costs tend to dictate things even in China.

For North America the quick wins are up and down the coasts offshore connecting the (coming) major GW-scale wind farms, and trans-continent to link the poorly connected grid zones. Personally I'd lay subsea on the Gulf coast and the Great Lakes to speed some of it up as well. NREL et al have all done the numbers on this many times, that is precisely why their study budgets get cut so often by the fossil lobby.

It is quite difficult to get costs down in this area. Globally demand for HVDC sets (etc) well outstrips supply, much more so than for Tesla BEVs !!! Change is coming, but the same factors that cause battery suppliers to hang back also cause HV manufacturers to be leery. Believe me, as I am in the discussions and I am one of the decision makers in my own small (but vital) piece of the supply chain. Boom and bust cycles are well known to them all, and the manufacturing facilities and associated supply chains are not cheap and not quick. This is one reason why small amounts of local battery and local solar can have more impact than macro-economic analysis would have one think at first, and so the changes that Tesla are making to Powerwall desgn & marketing/install are understandable in this context. (Though it is interesting watching Tesla try to impose a dominant design on the retail storage industry in a quarterly call with a screaming baby).

It is all about the batteries. This was the big risk still sitting in the middle of the table after this quarterly earnings call. It was addressed, discussed, but not assured as being solved. Tesla have taken the big decisions of building Berlin and Texas, and of ordering up the 4680 battery line equipment sets, whilst the actual Kato Rd 4680 pilot line is still not ready in all respects. There is hope that this will work out, and fear that it won't, and that is the biggest risk to both Tesla and humanity right now. They can get Berlin and Texas (and maybe storage) into a limping start without fully resolving that, but in order to scale as is required the 4680 dry process needs to be working at a high rate and good yield. Until that is happening Semis and Cybertrucks will just be unobtainable glamour models in small quantities.

It really is all about the batteries.

 

[Prunesquallor]

View attachment 657505

God, even with Tesla's help FCA/Honda are still near the bottom of the pack. These credits are going to be around a loooong time.

Weird that PSA-Opel show up in a separate pool. I thought I read that the Stellantis merger pulled them into the Tesla pool.

 

[Words of HABIT]

Dirty energy will soon have the same public perception as smoking.

Here here.

 

[MC3OZ]

I dare you to get a project dropping electric cables along populated coasts in Europe to be a quick project. Or major new ones over land for that matter.

Teslas issues with environmental permits for the Berlin factory wouldn't even register on that scale.

Actually. Wonder what the boring company would charge for a much smaller tunnel than they are doing now. Probably the cheapest. Almost certainly the fastest in many areas. Except there would be protests about damaging the groundwater etc.

Sigh, wonder how many solarpanels I can fit on my apartment balcony. I might have to soon.

I agree on smaller Boring co tunnels.

IMO Boring machines will eventually be made in a factory in higher volumes.

The current machine being the largest size, my guess is that there is scope for 2 smaller sizes. The smaller sizes should be easier to mass produce.

For in the sea HVDC I was thinking more of Australia, the US and Asia. However, Europe already uses this to connect Offshore Wind.

 

[MC3OZ]

Sure.

"Currently even though Tesla haven't solved FSD, Legacy auto makers have a hard time buying from Nvidia or other companies in scale due to the prohibited cost. Nvidia's L5 TRAINING platform is almost 10k, and this doesn't include the 8 or so lidars needed. Not to mention an entire cooling solution that need to dissipate about 3kw of power. Even so FSD isn't solved using this 30k-40k hardware package."​

Other companies chip doesn't use 3kw, you are using figures from 10 years ago. All self driving companies either use Nvidia, Mobileye, Qualcomm, Huawei or developed their own. Lets take for example NIO ET7 releasing Q1 2022. It uses 4x Nvidia Orin which provides 1,096 TOPs of power while using only ~180 watts. Each Orin chip is 258 Tops/45 watts.

This is compared to HW3's 144 Tops/ 100 watts, almost 8x more powerful with only 1.8 more wattage.
That's way more powerful and efficient than FSD Computer while also being smaller.

Now about the Lidars, its also a myth that Lidars are power hungry.

Livox 12 Watts
https://livoxtech.com/horizon/specs
Luminar 15 Watts
https://wired.com/story/lidar-cheap-make-self-driving-reality/…
Innovision under 40 Watts
https://oemoffhighway.com/electronics/sensors/product/21073236/innovusion-inc-innovusion-cheetah-lidar-system…
Innoviz One 15-23 Watts
https://just-auto.com/interview/ces-israeli-start-up-bags-capital-accelerates-lidar-innovation_id192792.aspx…

All listed lidar are high resolution & will go into mass production on cars this year & next year.
What about lidar costs? All the lidar listed above cost well under $500.

"Currently even though Tesla haven't solved FSD, Legacy auto makers have a hard time buying from Nvidia or other companies in scale due to the prohibited cost."​

To go even more into details not only will NIO deploy Nvidia, Volvo will also deploy 1x Orin in 2022 in their Volvo XC90.
2022 BMW iX will use 2x of Mobileye's next gen EyeQ5 (48 Tops/20 watts).
Geely's 2021 Zeekr 001 will use Mobileye's complete supervision (2x EyeQ5).
BAIC's 2021 Arcfox will use Huawei's ADS with their MDC 810 computer (400/800 TOPS).

I could go on and on.

​

​

"No auto company will copy this as they wouldn't even know where to start. Nvidia, Mobile eye, and Waymo might ditch their approach asap trying to catch up, which is a totally different strategy as they mainly focused on localization using Lidar and HD maps."​

The myth that Lidar and/or HD maps, slows you down or prevents you from scale needs to die.

There are two system's releasing by the end of 2021 that will have door to door city streets autonomy anywhere in china.
Remember that china is orders of magnitude harder to drive than in the US.

2021 Zeekr's 001 with Mobileye's Supervision
2021 Arcfox Alpha S with Huawei's ADS

Surely the issue with HD maps, is you need to maintain those maps and by defination HD is a lot of data, data that could change relatively frequently.

Also no one has every claimed that HD maps and Lidar mean you don't need to solve vision.

It is a shortcut, or to use Elon's term a crutch. So it does deliver a certain standard of solution quicker. The question is the impact of the timeframe for the total solution.

Rather than the shortcut Tesla is taking the long road and aiming for a total solution at the lowest cost with minimal on going maintenance.

If a company is behind and needs to get to market quickly a shortcut is the right option. Eventually the need to do it properly with a total solution which reuires fully solved vision.

HD maps and Lidar are more like a bet vision can't be solved quickly.

 

[Curt Renz]

ABC - this afternoon: Lordstown - Late Taxes

Excerpt:

The failure of an Ohio-based electric truck startup to pay $570,000 in real estate taxes due in early March is yet another troubling sign for a company that has been barraged by bad news this year.

Lordstown Motors Corp. stock has plummeted from nearly $31 a share on Feb. 11 to just over $10 on Tuesday in the wake of a U.S. Securities and Exchange Commission inquiry and the filing of four potential class-action lawsuits by investors who claim they have been defrauded.

 

[TheTalkingMule]

I really think we are going to be trading sideways for awhile yet. We are still growing into our current valuation and might be the rest of the year.

Normally I'd think the same, but the path is just way too stable now.

This period of consolidation and the "tightening wedge" is the normal TSLA rotation back into shorting. We're now seeing the short interest start to tick higher over the last few reports.

Back in the day there was likely to be some massive random hiccup that would make these short positions profitable. Nowadays......not so much.

Each earnings report should be considerably better than the last for the next 3. There's no opportunity to cover on the horizon, so I think it'll come to a head much faster.

Plus I think this quarter's deliveries really caught MM's off guard. They were counting on a miss and their synthetic short positions are likely at the higher end of acceptable. When the hell do you cover that?

There aren't many buyers, but there's a hell of a lot of shorters and no sellers. Someone's gotta jump first eventually!

 

[Singuy]

Sure.

"Currently even though Tesla haven't solved FSD, Legacy auto makers have a hard time buying from Nvidia or other companies in scale due to the prohibited cost. Nvidia's L5 TRAINING platform is almost 10k, and this doesn't include the 8 or so lidars needed. Not to mention an entire cooling solution that need to dissipate about 3kw of power. Even so FSD isn't solved using this 30k-40k hardware package."​

Other companies chip doesn't use 3kw, you are using figures from 10 years ago. All self driving companies either use Nvidia, Mobileye, Qualcomm, Huawei or developed their own. Lets take for example NIO ET7 releasing Q1 2022. It uses 4x Nvidia Orin which provides 1,096 TOPs of power while using only ~180 watts. Each Orin chip is 258 Tops/45 watts.

This is compared to HW3's 144 Tops/ 100 watts, almost 8x more powerful with only 1.8 more wattage.
That's way more powerful and efficient than FSD Computer while also being smaller.

Now about the Lidars, its also a myth that Lidars are power hungry.

Livox 12 Watts
https://livoxtech.com/horizon/specs
Luminar 15 Watts
https://wired.com/story/lidar-cheap-make-self-driving-reality/…
Innovision under 40 Watts
https://oemoffhighway.com/electronics/sensors/product/21073236/innovusion-inc-innovusion-cheetah-lidar-system…
Innoviz One 15-23 Watts
https://just-auto.com/interview/ces-israeli-start-up-bags-capital-accelerates-lidar-innovation_id192792.aspx…

All listed lidar are high resolution & will go into mass production on cars this year & next year.
What about lidar costs? All the lidar listed above cost well under $500.

"Currently even though Tesla haven't solved FSD, Legacy auto makers have a hard time buying from Nvidia or other companies in scale due to the prohibited cost."​

To go even more into details not only will NIO deploy Nvidia, Volvo will also deploy 1x Orin in 2022 in their Volvo XC90.
2022 BMW iX will use 2x of Mobileye's next gen EyeQ5 (48 Tops/20 watts).
Geely's 2021 Zeekr 001 will use Mobileye's complete supervision (2x EyeQ5).
BAIC's 2021 Arcfox will use Huawei's ADS with their MDC 810 computer (400/800 TOPS).

I could go on and on.

​

​

"No auto company will copy this as they wouldn't even know where to start. Nvidia, Mobile eye, and Waymo might ditch their approach asap trying to catch up, which is a totally different strategy as they mainly focused on localization using Lidar and HD maps."​

The myth that Lidar and/or HD maps, slows you down or prevents you from scale needs to die.

There are two system's releasing by the end of 2021 that will have door to door city streets autonomy anywhere in china.
Remember that china is orders of magnitude harder to drive than in the US.

2021 Zeekr's 001 with Mobileye's Supervision
2021 Arcfox Alpha S with Huawei's ADS

I don't see any prices listed.

Can you provide the price of 4 Orin chips and say 4 lidar for 2022?

The point is scalability which means cost. One lidar being 500 bucks each and a few Orin chips being 2-5k doesn't exactly scream..I can put these things on a 25k car. Teslas solution cost 500 bucks total in 2018...

 

[EVWatcher]

I think the unknown details of Biden's tax plan have investors on edge and it's supposed to be unveiled tomorrow. I know many have stated that this forum has plenty of non Americans here but its worth keeping in mind that people on are edge and some are pulling their money in anticipation of this new bill which drives the price down whether you like it or not. I myself had a wealth manager implore me sell because of the Biden tax plan - I told him I was not interested in his services. Given that the plan could be retro active to Jan 1, 2021 or kick in Jan 1, 2022 or undergo pretty heavy changes between its current form and when/if it gets enacted I think the fear of uncertainty could push the the price down until at least the details of the proposal are made public, which as I understand is tomorrow night. Keep in mind there are lot's of newly minted investors taking the advice of advisors like above.

A note about the call. Did nobody else get the feeling when Musk said electricity usage would triple, he then caught him self, muttered an expletive as if he said something he didn't mean to say, and then said doubled for EV's and then circled back to triple for HVAC. Struck me as telling as to whats on his mind.

 

[TheTalkingMule]

I myself had a wealth manager implore me sell because of the Biden tax plan - I told him I was not interested in his services.

Wait....what? Was he worried about growth stocks in general or Tesla's business being negatively impacted? If it's the latter you should really offer to pay for a CATScan. People are so bizarre.

 

[StarFoxisDown!]

Normally I'd think the same, but the path is just way too stable now.

This period of consolidation and the "tightening wedge" is the normal TSLA rotation back into shorting. We're now seeing the short interest start to tick higher over the last few reports.

Back in the day there was likely to be some massive random hiccup that would make these short positions profitable. Nowadays......not so much.

Each earnings report should be considerably better than the last for the next 3. There's no opportunity to cover on the horizon, so I think it'll come to a head much faster.

Plus I think this quarter's deliveries really caught MM's off guard. They were counting on a miss and their synthetic short positions are likely at the higher end of acceptable. When the hell do you cover that?

There aren't many buyers, but there's a hell of a lot of shorters and no sellers. Someone's gotta jump first eventually!

Pretty much feel the same. I think they have 2 months to short/cap the stock, but they're gonna have a hard time dropping it significantly lower than this level because if they tried to drop it into the 500's again, even old school investors would start eyeing that P/E (If it dropped to 500's and Tesla posts around a $1-1.25 EPS for Q2 the P/E would drop to like 200 or even lower).

I still think Shorts and MM's wouldn't dare be caught manipulating as the Biden Infrastructure deal nears ( July timeframe). Way too high of a risk of repeat of one of the moves in 2020 and they would be obliterated. They're still playing with fire when it comes to V9 and subscription model. We know it won't drop in April (and likely not May), but it could drop anytime in June.

 

[EVWatcher]

Wait....what? Was he worried about growth stocks in general or Tesla's business being negatively impacted? If it's the latter you should really offer to pay for a CATScan. People are so bizarre.

There were a number of things that concerned me about his honesty and or grasp on the market. However one thing that did make sense is that there are a lot of people who made a lot of money in Tesla in the last year and they will panic and want to lock in their gains before a tax hike. It occurred to me that that would only be a concern for people getting advice from people like him - keep in mind people like that are paid for managing your investments, and they can't manage/get paid for an investment they themselves didn't make -, however I just politely told him I wasn't interested in his services and I would let him know if I changed my mind.

 

[StarFoxisDown!]

There were a number of things that concerned me about his honesty and or grasp on the market. However one thing that did make sense is that there are a lot of people who made a lot of money in Tesla in the last year and they will panic and want to lock in their gains before a tax hike. It occurred to me that that would only be a concern for people getting advice from people like him, however I just politely told him I wasn't interested in his services and I would let him know if I changed my mind.

That already pretty much happened. We went through a 40% sell off and consolidation period as well as now past the weakest quarter of the year with growth ahead. Everyone that was worried about their gains sold. If the stock never had that 40% correction, then sure. But that's terrible advice now

 

[EVWatcher]

That already pretty much happened. We went through a 40% sell off and consolidation period as well as now past the weakest quarter of the year with growth ahead. Everyone that was worried about their gains sold. If the stock never had that 40% correction, then sure. But that's terrible advice now

I think there is the x-factor of people who are blindsided by Biden's tax plan announcement from last week. I think there are probably a lot more very novice investors in TSLA than people tend to think.

I agree his advice was terrible tho. This is advice he gave me after I pointed out the tax hike was in part going to fund a lot of programs Tesla would benefit from. He provided me a nice excel table showing what the price of Tesla would need to be post tax hike to equate to what it was before the tax hike. He also went on a diatribe about Cisco. It was both entertaining and frustrating.

I'm just saying not every Tesla investor in existence is looking at this forum or has a good grasp on true long term investing.

 

[StarFoxisDown!]

I think there is the x-factor of people who are blindsided by Biden's tax plan announcement from last week. I think there are probably a lot more very novice investors in TSLA than people tend to think. I agree his advice was terrible tho. I'm just saying not every Tesla investor in existence is looking at this forum or has a good grasp on true long term investing.

Definitely true that there are novice investors in TSLA that would likely have a emotional reaction to even the smallest of things because that are uninformed

But as I've mentioned before, the effect of retail traders is dramatically overexaggerated. They could all sell out of their TSLA position and it would only move the stock a couple of percentage points. The big boys control the shares that are available while a vast majority of the shares are locked up by insiders (and now S&P funds)

 

[Singuy]

I wouldn't dare to make any huge sells until the tax plan is well understood. It could be retroacted and those who made large sales can end up screwing themselves badly.

 

[StarFoxisDown!]

I wouldn't dare to make any huge sells until the tax plan is well understood. It could be retroacted and those who made large sales can end up screwing themselves badly.

Also I think people are getting the 2 Biden plans confused. The Infrastructure bill and the tax increase on corporations is one which is likely happening in July timeframe...while a 2nd individual stimulus bill is being discussed that has the capital gains tax increase associated with it. That one likely won't even pass this year (And I think they'll definitely do a 1 or 2 year delay on that tax increase taking effect)

 

[goinfraftw]

It's $3500 pre-split right now. A year ago, it was $800. 5 years ago, it was $240. Let that bake in.

 

[Prunesquallor]

https://twitter.com/x/status/1387176157737541641

https://twitter.com/x/status/1387180114836967427