Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

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

This site may earn commission on affiliate links.
In addition, I would say that it was a good example of 7 vehicles that either don't have AEB or the AEB system they have isn't very good. (I think it looked like some of them slowed down and had minimal impact, but some of them didn't appear to slow down at all.)

Assuming that there wasn't a mechanical/electrical failure of the lead Tesla there were 8 drivers in a row "asleep" at the wheel.

I probably enjoy driving more than most people and I wouldn't particularly enjoy seeing that privilege taken away from humans but, my, oh-my, that video is the perfect example of why humans probably shouldn't be allowed to drive cars anymore. It really shows how poor of drivers most humans really are.

I mean, how hard can it be to look ahead and know when to brake?
 
Interesting white paper on the Tesla Solar Inverter posted on Reddit today: https://tesla-cdn.thron.com/static/...tecture_White_Paper_NA-EN_12212022_PYC2C9.pdf

Some key points via the comments:
  • $1,400 to $1,800 cheaper upfront cost compared to the alternatives
  • 6-15% cheaper in terms of $/kWh during the lifetime
  • 38% more reliable
  • Scalability: Easy to install and service + less power electronics per solar installation
  • Supply scalability: 1.4 TW of in-house built inverters (vehicles + energy) already delivered (33x of SolarEdge and Enphase combined)
Aside from the above, the most interesting thing to me is that Tesla seems to be positioning this to be a product they provide to third-party installers. From the conclusion: "We look forward to extending these advantages to installers and system owners across the world to accelerate the transition to a sustainable future."

Selling turn-key solar systems to third-party installers really opens up Tesla Solar's total addressable market.

Edit: Source's source confirms they're shipping now: "Our residential solar inverter is shipping now to Tesla Certified Installers"
 
Last edited:
And the stamping presses do need those fancy isolation foundations.

... yet another advantage to CT-style exoskeleton: no stamping press needed. Even some exotic die casting process is a candidate at this point. But I don't see Tesla holding back growth for last-Century manufacturing tech like stamping. ;)

1936 Ford Tudor in SS.jpg


Yeah, the trick is going to be locating the factory close to the nickel mine. That's the critical path: it's all about LOGISTICS!

Cheers!
 
FYI this guy Will has leveraged Truflation's "real time" estimate of YoY inflation into predictions for what the Fed CPI print will be. He can do this because of the lag in the Fed's data.

He has been pretty accurate for the last few prints I believe.

He is predicting a print of 6.8%, 0.3% more than forecasts. FYI

 
Last edited:
It's not clear to me what the ultimate Austin and Berlin capacities are for Phase 1. 500k apiece? 1 million?

Let's use Giga Shanghai as our base case for minimun capacity. That's 3 lines, total about 1.1M current capacity. So let's call it 400K/yr/line in the next gen factories (Berlin, Austin). So both Austin and Berlin will have 2 lines each, that'd be ~800K/yr each in the base case.

I expect Berlin will get an new line for "Model Zwei" next, and we expect 2x capacity in the same space. So 1.6M/yr there, but will probably require an new building, so subject to the vagaries of the German permitting process (maybe helpful if France invaded?)

Giga Austin will ramp quickly, depends more on the new Cathode plant, then the new LiOH refinery in Corpus Christie (expected to come online in 2024). It will be huge. I expect Giga Mexico will get it's LiOH from Corpus Christie too.
 
Interesting white paper on the Tesla Solar Inverter posted on Reddit today: https://tesla-cdn.thron.com/static/...tecture_White_Paper_NA-EN_12212022_PYC2C9.pdf

Some key points via the comments:
  • $1,400 to $1,800 cheaper upfront cost compared to the alternatives
  • 6-15% cheaper in terms of $/kWh during the lifetime
  • 38% more reliable
  • Scalability: Easy to install and service + less power electronics per solar installation
  • Supply scalability: 1.4 TW of in-house built inverters (vehicles + energy) already delivered (33x of SolarEdge and Enphase combined)
Aside from the above, the most interesting thing to me is that Tesla seems to be positioning this to be a product they provide to third-party installers. From the conclusion: "We look forward to extending these advantages to installers and system owners across the world to accelerate the transition to a sustainable future."

Selling turn-key solar systems to third-party installers really opens up Tesla Solar's total addressable market.

Edit: Source's source confirms they're shipping now: "Our residential solar inverter is shipping now to Tesla Certified Installers"
More summary:
  • Tesla is vertically integrating more by leveraging solar panel fleet data and vehicle inverter design and manufacturing experienceto make a custom inverter design for rooftop solar
    • 1.4 Terawatts of inverters made when including vehicles; 33x more than Enphase and SolarEdge combined
"We specialize in power electronics, and we harnessed our expertise to produce a reliable solar inverter."

  • Tesla arguing for elimination of microinverters and optimizers for most roofs and return to traditional string inverters, a reversal solar industry trends in recent years.
    • Minimum-cost solution is now to minimize power electronics costs at the expense of a bit of overall energy output.

This higher cost [of module-level power electronics] may have been justifiable when equipment costs were 2-3 times what they are today. However, as PV module and other solar equipment costs have dropped, inverter costs have become a larger portion of the total system cost and a more impactful driver of LCOE [Levelized Cost of Energy].
  • Improvements on traditional string inverters
    • More MPPTs (maximum power point trackers)
    • Accepting lower input voltage per MPPT, meaning high efficiency maintained with as few as two panels per string
    • Improved reliability
    • Software integration with overall Tesla energy system (powerwall, vehicle, panels, maybe HVAC someday)

1673470527545.png


1673470895083.png



If there is an installation error, Tesla Solar Inverter has built-in cellular, Wi-Fi, and Ethernet connectivity and is easy to inspect and repair with a single unit that is accessible from the ground. In comparison, if there is a wiring or communication issue on a single microinverter or optimizer, professionals must climb on the roof and remove modules to troubleshoot and perform remediation, a physically challenging and dangerous process. If the system includes energy storage, installing Powerwall+ instead of a Powerwall 2 with a third party inverter reduces the number of unique components on site and moves all commissioning activities into a single app.
 
Last edited:
I don't know where to start on this, I don't have much to take issue with the final conclusions because they are so broad and general and long-term, they are nearly meaningless and it seems a lot of very specific and very recent data was mined to try to a broad understanding how this might play out over time. While I applaud the discussion, I think it's a misguided way to come to the broad, general conclusions because we know there is always a cost to growth. That is best modelled in from a general case without using specific data from limited timeframes without controlling for other variables and a general lumpiness of data.

One example, the circled data in the table above represents the automotive GM% difference between Q1 and Q2 2022 of 5%. However, you are using gross margins including regulatory credits which should be backed out of such an analysis due to their high variability. Tesla recorded $344 million in regulatory credits for Q2 2022, down 49% from Q1 which was $679 million. That's a loss of around $1,200 per vehicle due to quarterly regulatory credit variation which is essentially random (not correlated to quarterly sales at all). Backing this out would have the effect of reducing the 5% drag to GM you attribute to ramp costs by over 20%.

Another example is that Q2 had a rather precipitous drop of over 15% in production over Q1 due to COVID shutdowns. That naturally had a strong downward pressure on GM's that is seperate from the drag of ramp on GM's. We don't have enough data to quantify which impacts contributed how much. But I would guess COVID shutdowns had an even stronger impact on GM's than that of not backing out regulatory credits:
View attachment 894521

If sharply lower regulatory credits, disrupted production due to COVID shutdowns and costs of ramping all had negative impacts on GM's in Q2, but they only fell 5%, then it looks to me like those impacts might be hiding some very significant gains in terms of manufacturing efficiency (in addition to the higher selling prices and potentially other hidden positive impacts).

Using the quarter over quarter GM decline between Q1 and Q2 2022 to estimate the impact of ramping on GM's is simply not possible and a misguided use of the data we have. Stepping back and looking at the big picture is almost always more productive than trying to extract meaning by using a magnifying glass.

One large variable in terms of overall margin potential, which you correctly state is not knowable, is margin compression due to pricing pressures which would either be due to market conditions (most likely, IMO) or competitive pressures from other manufacturers (very unlikely, IMO). Either possibility can be potentially offset by further increases in production efficiency. I think this, combined with changes in market prices, is the most important variable when speaking of long-term margins and any attempt to quantify long-term margins without discussing production efficiency is suspect.

You make no attempt to quantify how production efficiency could impact Tesla's competitiveness going forward and that is the same mistake many analysts make when they just assume all manufacturers are roughly equal in their ability to manufacture efficiently and to constantly improve that efficiency. That it must cost the same to produce a product that is comparable in functionality to a competitor's product. But Tesla's constant production innovations and optimizations have been blowing that misguided notion wide open (at least for those willing to look at the trends). That's why Tesla's current margins are so much higher than competitor's margins.

Ignoring cost to produce for a moment, Tesla's current margins are not really limited by what other manufacturer's sell their products at, margins are limited by how much people are willing to pay, how much they can afford in any given economic climate. Because there are not enough cars available of the kind people want to buy in the price range they are willing to pay. As the cost to produce continues to decline and volumes continue to rise I will suggest that Tesla's biggest competitor will continue to be themselves as they ramp volumes ever higher.

All manufacturers are not approximately equal, and this is not a static dynamic, Tesla's lead in terms of cost to produce is slowly increasing. Look at the trend of Tesla's cost to produce each car over time. I would suggest this is not entirely attributable to volume efficiencies that have been realized so far, and that will continue to accrue, but also to a superior way of looking at every challenge and the desire to take on those challenges in an efficient manner.

Elon has made no secret that he has big goals in terms of increasing speed of production throughput as well as continuing to increase the density of manufacturing in three dimensions. The reason Tesla strives for volumetric production efficiencies is because this leads directly to a reduction in the investment required for building new production facilities. This plays directly into the cost to continue ramping production and is a big deal as Tesla ramps production from 2 million vehicles/year to 20 million/year.

Not only does a smaller factory cost less to build, but more cars per day from a smaller footprint has compounding returns in efficiency. A smaller plant requires less time for people and materials to move about. A smaller plant requires less land area to clear and build (which further reduces capital expenses) and improves efficiency of movement outside the building. Any competitor who thinks Elon's next factories will be cut/paste copies of existing factories are in for a rude awakening because Elon did not go into automaking to be like the rest. Every generation is a big improvement, Elon does not believe in limitations beyond those imposed by physics itself as guided by economics. These efficiencies manifest slowly, as each new generation of plant is optimized, and production gradually increases.

The future is always hard to see, and there will be surprises along the way, some positive, some negative. But the overall trend is clear and without seeing a reason why that trend is coming to an end, I will continue to bet on Elon and his teams to continue to improve at a rate faster than the rest. The Chinese will be the closest competitors, but I don't think even they will be able to match the efficiency of the systems Tesla is implementing. Chinese manufacturing rides on Elon's coat tails which is why they invited him in to begin with. The Chinese are not infallible, they are just smart, focused and willing to work hard. If they had the vision to exceed what Elon is doing, they wouldn't have invited him in, they follow in his footsteps. At best they can match it a few years later. Never say never but the trend is your friend and I have not seen that the trend is not favorable.

100% agree with everything here (incl why China gave Tesla a pass on 50% JV ownership). The only potential storm cloud I see is the ability of battery materials to ramp supply fast enough for Tesla. And Tesla knows this too of course, which is why they are cutting mining deals and building lithium refining plants. Us investors haven’t been given the information needed to know whether Tesla has a clear path to 50% YoY growth from here on out WRT being to acquire enough battery cells and minerals to make such. I am hoping Tesla will provide that info on March 1st.

Market demand will be there as Tesla brings new products online. EV competition won’t be a factor since no one else can match Tesla cost and margins. Non battery materials will be there as Tesla displaces gas powered cars. All this assumes Tesla executes well.

Onwards and upwards.
 
Last edited:
Cartographic specialist here breaking silence once again to rap knuckles. Fortunately, class today involves maps, so…..

Before knuckle-rapping begins, a word for all to remember: any time someone shows especially a full-global map, ask yourself if that is the most appropriate projection of a spheroid’s surface onto a flat plane.
For this situation, in order not to be mischievous, it must be an equal-area projection; the one shown most decidedly is not. Rather, it shows off the desired focus in the best possible light. Below please find my choice; unfortunately, I am not going to place the circle on it but leave it to each of you to do so - use as circumference markers the west boundary of Pakistan, the northern boundary of Hokkaido, and the southern extremities of Indonesia (and on edit: it is a circle of radius 4100km, centered on Guiyou, Guizhang Prov.).
You will note that this region covers a far greater fraction of the earth’s land surface than appeared to have been the case in the original version.

Link here: http://www.equal-earth.com/Equal-Earth-Map-150E.jpg

Final grade: a gentleman’s C-.
Hmmm, didn't know you were a Cartographer. (When people ask me what a Cryptographer does, I tell them I take pictures of cemeteries.) When the Fukushima earthquake happened, Qualcomm's Sydney AU office got calls from the IT people asking if we were OK. I had to point out to them that San Diego was actually closer, it just didn't look like it on the usual maps. Not close in either case, but...
 
Then to be clear, without the second stamping portion that is under construction, capacity is 250k?

Fremont only has one large Schuler press to make all it's Models 3/Y combined. That's close to ~12K/week now?

Tesla Stamping and Assembly of Model 3 | The Future is Forming | 2018/09/20

I think it's more about the dies than the press. Tesla has also increased the capacity of the Fremont press by over-clocking it compared to a stock press (typical cowboys).

Cheers!
 
The stamping area that I saw when I visited Giga Berlin a year ago is for 500k vehicles. It's darn fast and the even store and buffers finished stampings. That's up and running for a longer time.
OK, then the new stamping section on Phase 1 is for an additional 500k vehicles, i.e. a total one million?

This is the origin of my wondering. In the Summer or Autumn, the stamping in Phase 1 seems like it will be roughly double the size needed for 500k vehicles. @Artful Dodger's calculations seem to track with the footprint being built.
 
Roadster deliveries in 2023?

Given the lack of updates about status of Roadster deliveries and also the fact that I could be easily be earning 3.5% on my deposit money, I just called to ask for my deposit back. Took 20 minutes to get someone on the line; after taking my info and then having some muffled conversations with people in the background, the person came back and said "I'm being told that Elon has said we will deliver these in Q2 or Q3 of this year". I asked her to verify that she meant 2023. More muffled convo in background, and then she says "yes, in 2023".

I'm not sure how confident to be with this info, but I told her I would hold off on cancelling for a few months and see where this goes.

Personally I haven't seen anything on Twitter or elsewhere to support this; seems like most people were thinking 2024 for Roadster deliveries to start.
Hmmmm if they see inventory growing in the Model 3s and Ys right now and they have existing orders for Roadsters to fulfil, maybe it makes sense to shift focus to those? Maybe they've really been secretly ready to pull the trigger for a while now but needed to focus on satisfying 3/Y demand and are prioritizing based on order backlog?

I mean food for thought although this might not be overly appropriate for the roadster, but Tesla had recently requested an extension of confidentiality with the FCC around a new HD Radar unit because they were planning to market it on something in mid-January aka any day now.
 
Hmmmm if they see inventory growing in the Model 3s and Ys right now and they have existing orders for Roadsters to fulfil, maybe it makes sense to shift focus to those? Maybe they've really been secretly ready to pull the trigger for a while now but needed to focus on satisfying 3/Y demand and are prioritizing based on order backlog?

I mean food for thought although this might not be overly appropriate for the roadster, but Tesla had recently requested an extension of confidentiality with the FCC around a new HD Radar unit because they were planning to market it on something in mid-January aka any day now.

The Roadster is a halo car. It will never sell in enough volume to make a meaningful difference in overall corporate profits.
 
The Roadster is a halo car. It will never sell in enough volume to make a meaningful difference in overall corporate profits.
Maybe, but maybe now is just the first good time in recent years to start pumping them out

I think people were speculating the HD Radar is for Highland. But if we're talking Semi, CT, Highland, AND the roadster all suddenly in basically the same year, that's a crazy year.