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.
So, why, do folks think, that all those Model Ys at Giga Texas won't be delivered until April 7th?

Ready, set, go!

(I don't know why)
Mainly comes down to does Tesla want better gross margin or better operating margin in Q1

Better gross margin - wait until April 1st to start deliveries

Better operating margin - deliver the cars as soon as possible
 
Tesla Daily (Rob Maurer) which is quite good for straight news and stock info.
Dave Lee on Investing which is hit or miss. Some truly fantastic stuff mixed with some very blah/ stuff. I keep the skip button handy and have no trouble just skipping whole episodes.
Ride the Lightning is passable. Sometimes the call-in sections are interesting, it’s almost like “Tesla for beginners”, but has some good stuff mixed in.

I’ve tried a few others, but none of them stuck. I will say I do occassionally listen to the Electrek podcast. The daily one which is just a quick rehash of news is ok. The longer one is mostly meh.
I agree with all that, and I would add the Tesla Economist YouTube channel to that list. He can occasionally be overzealous; but generally carefully researched and efficiently presented, accurate information.

The obvious oversight, though, is TMC investor forum FTW. What an amazing smorgasbord of experts in various fields. Sure the occasional hothead or carebear commands an ‘ignore’✔️. But this place is truly a blessing for the patient and open minded Tesla investor.

As our HODL’ing skills are being (again) tested I feel a word of gratitude to you beautiful degenerates is appropriate.
 
In return for my not deleting normally-quashed posts about Lidar, here is some weekend humor. From the archives, a snap during a trip to Arizona in winter 2018. Shows priorities?

View attachment 779992
Weekend OT and Ukraine break c'ed, esp for those not cognizant of Tesla and Space X's history - this is a Klasse encounter, 10 year anniversary of sorts :

In defense praise of Mercedes Benz: Tesla (and Space X) might have been beached back in '08 had Mercedes not rescued them from the brink of bankruptcy by buying $50M worth of TSLA, and commissioned Tesla for the electric power train of their BEV Model B.

Also: like the '18 Model X, this '08 Mercedes E 350 is a "classic", not in the sense that it is over 25 years old, but its M272 E35 V6 engine is the last of its generation, completely perfected & debugged like the whole car actually. If you bought a 2008-9 E350 or its sibling the CLK350 you had a car with ZERO defects and super minor recalls (a solenoid in the fuel pump that could create problems in case of accident, bulletin 2016, or some latex deteriorating in some models threatening sunroof separation, recall bulletin ca 2021)

How do I know, uh I bought a '08 CLK350 three years ago (at 28,000mi) on a whim and ferreted out every possible detail about it back then, after purchasing it on instinct, just to be sure (2). Which incidentally led me to "discover" that Tesla wasn't some crazy rich billionaire toy as I had been led to believe back then (one has only so much time to learn about anything, we really depend a lot on our tribe's beliefs).

The lucky ending - since my TSLA investments paid off big time (1), why not indulge even if I really don't need a car, living in NYC, where the Benz stays 95% of its time in a garage. Plus the Y is mature enough to be at a good stage of its evolution; the next stage would be a couple years down the road, when all the kinks and improvements to the 4680 model will have been figured out.

(1) Trades, actually, some of which made the pot spectacularly big, then small.. -> lesson painfully learned, I would indeed have been better off just sitting on my TSLA stonk

(2) yeah, I'm sort of good with some tech stuff - decades of computer work ... hah ha even w/ mainframes, debugging software, you could sometimes figure out something's wrong with the algos when you felt the hardware shaking slightly
 
  • Like
Reactions: Discoducky
That’s why when any autonomous solution company says they want to do “sensor fusion” , it’s a red flag in my book. Many people don’t seem to understand.
I might be a bit biased as my master was in Sensor Fusion and my thesis was about Lidars. And first I hate semantic discussion, but I just want to clarify that doing sensor fusion can mean many things and it’s not black and white. When Tesla feed the odometer into their neural network, technically they are doing sensor fusion. Some might even argue that using multiple cameras is sensor fusion.

The approaches that differ is that some companies want to make two individual systems for perception, one camera based and one lidar based. Both will pick up some signals and miss some. Then they can choose to either use the one that is most certain or be safe and just use the one that is least aggressive, just break for every object either of them sees etc. Or both could be feed into a neural network and let it decide how to do the sensor fusion.

Then the question is if a map should be considered a sensor or not. Imo it should. It’s like any sensor, it has some signal and some noise/limitations. Just different properties in what it can detect and it tend to fail temporally which is unlike many other live sensors, and it needs for method for positioning the ego vehicle in the map. And then there is the wide spectrum of map represenations from very limited list of waypoints GPS coordinates to HD maps of Lidar Point clouds and all possible abstractions inbetween. And should the fusion happen inside a neural network or by some clever code. And if you are gonna be training a neural network on map data, then the question is how you deal with updates in map data and map formats over time in your training data. It adds some complexity, not impossible to solve but some complexity you need to deal with.

And there is also the difference in running the sensor fusion offline to generate labels or running it online to be used for control of the vehicle in the real world. Imo taking away that tool from the label team seems a bit silly.

Imo I think Elon is thinking correctly about sensor fusion. Like he says, deep learning is something you only use because you have to. And often less is more, removing parts, sensors, laws etc is often the best solution. And the most important with deep learning seems to be to get a big, diverse, high quality dataset. Tesla will soon update their cameras which will give them two sets of sensors in their dataset. That will be tricky enough, how many sets of sensors will waymo have? How can they statistically prove anything with so many different sets of sensors?
 
Last edited:
Tesla is now prioritizing orders that include FSD. I bet they will do the same for the new Model Y Standard Range AWD that will most likely be produced in Austin with 4680 cells.

They won’t be able to make many of these due to ramping of 4680s (these cars will not get LFP), but demand will be thru the roof on what will likely be an ~$50k base AWD Y.

Best way to manage demand while also maximize margins will be to deliver to those who add FSD. This is gonna be interesting to see how it plays out.

(I’m bummed as the odds that my MYP or MYLR order will come from Austin with 4680 is now low. This SR AWD w 4680 is the obvious move for Tesla.)
 
Tesla is now prioritizing orders that include FSD. I bet they will do the same for the new Model Y Standard Range AWD that will most likely be produced in Austin with 4680 cells.

They won’t be able to make many of these due to ramping of 4680s (these cars will not get LFP), but demand will be thru the roof on what will likely be an ~$50k base AWD Y.

Best way to manage demand while also maximize margins will be to deliver to those who add FSD. This is gonna be interesting to see how it plays out.

(I’m bummed as the odds that my MYP or MYLR order will come from Austin with 4680 is now low. This SR AWD w 4680 is the obvious move for Tesla.)
These 2 new “Shorter Range” Teslas are super intriguing to me. Well at least if they are 4680/ Austin based.

On the surface, you’d think they would be worse for road tripping, but the smaller cells, higher efficiency, and likely higher charge rates should make them charge exceptionally fast. The new 244 mile range Model Y has a 26% shorter range, but a 33% smaller battery pack. It’s very likely it will charge at 250 kW for the first 50% of the charge cycle (based on Musks comments which suggested the 4680 cells would be “Next Level” compared to the Model S). You might have an extra Supercharger stop on a longer trip, but all over your Supercharger stops will likely be around 10-15 minutes instead of 15-20 minutes.

The “Model Y AWD” isn’t as efficient as it’s SR cousin, but should share much of the charging speed advantages. Enough so that I might well be willing to sacrifice the extra range to get that charging speed advantage. Even if it’s not much cheaper than the LR model.

Obviously if these guys are LFP, it’s a whole other set of compromises.
 
The new 244 mile range Model Y has a 26% shorter range, but a 33% smaller battery pack. It’s very likely it will charge at 250 kW for the first 50% of the charge cycle (based on Musks comments which suggested the 4680 cells would be “Next Level” compared to the Model S). You might have an extra Supercharger stop on a longer trip, but all over your Supercharger stops will likely be around 10-15 minutes instead of 15-20 minutes.

Obviously if these guys are LFP, it’s a whole other set of compromises.
Yes. It may also be chargeable up to 100%... we have to see. Also it may get to see the new max charging power of 324kW that has been mentioned recently. Imagine 324kW for a significant portion of your SOC range.

As I've said before I believe the 4680 will be the step change that eliminates "short refueling times" as one of the heretofore unbeatable advantages of ICE. The length of a trip to a gas station will more-or-less equal the length of a Supercharger+4680 charging session... or if one is longer than the other it isn't worth complaining about. Especially as Tesla deploys Supercharger sites that deliver 324kW or higher. Will be nice to get faster and faster charging rates, years into ownership of a 4680 vehicle. I don't think Tesla have truly determined what rate they will stop at.

(For comparison my car briefly hits 120kW if I'm lucky and quickly falls down to the 80 and 70kW zones. Oh the humanity! 🤣)
 
I only care about Y from A factory as an investor. Having reports of A deliveries will be good, make me happy.

As an owner, and a driver, I really care not. It is a bit akin to what grade of gas one puts in an ICE, or even diesel vs gas. A non issue as an owner, just a fact, not something that improves the ownership experience or happiness.

I think it is potentially a thing of beauty to consider how holdouts for A cars might actually be subject to FUD from some clever source, trying to delay or cancel orders, push them back, allow survival of other brands. Conspiricists here abound, and subtly the bulls are the FUDsters.

I am pretty sure these same Y waiters will become Yners and complain about panel gaps, fit finish, road noise, missing parts, etc.

I wish everyone that gave a dang about founders storylines had something better to do with their time.

Complaints about the TMC blog or podcast, or whatever, same to u. Shut up and go do it yourself. Everyone else already has.
 
Timing. Simple as that. You can't redefine a term. The company was founded before Elon was involved, he came along about 8 months later.

Tesla before Elon DOES NOT EQUAL Tesla after Elon, they are two VERY different companies. He created the company which we invest in today, and without him it would have never existed. Tesla would have surely failed without Elon, heck it almost failed WITH him.

There are suggestions that Tesla should make an extra long range Model Y with 4680s, which seems possible, but I think it's highly unlikely they do that.

I think it's very possible Tesla will increase the ranges of all their models across the lineup, especially if they ever migrate everything to 4680 cells, which I think is extremely likely a few years down the road. Once 4680's get into ludicrously high production volume it will become super cheap and easy for Tesla to throw a few more cells into each battery pack to ensure they lead the industry in range.
 
Last edited:
These 2 new “Shorter Range” Teslas are super intriguing to me. Well at least if they are 4680/ Austin based.

On the surface, you’d think they would be worse for road tripping, but the smaller cells, higher efficiency, and likely higher charge rates should make them charge exceptionally fast. The new 244 mile range Model Y has a 26% shorter range, but a 33% smaller battery pack. It’s very likely it will charge at 250 kW for the first 50% of the charge cycle (based on Musks comments which suggested the 4680 cells would be “Next Level” compared to the Model S). You might have an extra Supercharger stop on a longer trip, but all over your Supercharger stops will likely be around 10-15 minutes instead of 15-20 minutes.
This might be correct if it weren't for the massive gaps between Superchargers in the central part of North America--particularly if you are traveling north-south rather than east-west. One Supercharger location out means that you had better have a long range Tesla to have a chance of making it to the next Supercharger. The FindUs page doesn't show any near term improvements, so it will be years before this plan would have a chance of working.
 
This might be correct if it weren't for the massive gaps between Superchargers in the central part of North America--particularly if you are traveling north-south rather than east-west. One Supercharger location out means that you had better have a long range Tesla to have a chance of making it to the next Supercharger. The FindUs page doesn't show any near term improvements, so it will be years before this plan would have a chance of working.


Gaps have been filled in pretty nicely with continued aggressive supercharger expansion. Where specifically are you referring?
 
These 2 new “Shorter Range” Teslas are super intriguing to me. Well at least if they are 4680/ Austin based.

On the surface, you’d think they would be worse for road tripping, but the smaller cells, higher efficiency, and likely higher charge rates should make them charge exceptionally fast.

In what universe is 4680 a smaller cell than 2170?

edit: Maybe you meant "less populated module", I say that because "smaller pack" would be wrong also. The pack is physically the same size.
 
Gaps have been filled in pretty nicely with continued aggressive supercharger expansion. Where specifically are you referring?
OK, KS, NE, SD, ND, etc. Try going north-south with even one of the Supercharger locations out of action (Especially Perry or Salina, both of which have been down on occasion--I know, I got stuck there). East-west isn't too bad, but North-South needs LR+ for normal unworried travel. Double this in winter, or with an older Tesla.