-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
- Status
SunCatcher
Member
Investor Almanac article from 3 days ago. Musk's message holds; SpaceX will stay private, and for good reason.
larmor
Active Member
Wow... Even if it is old, it gives a huge insight to how a growth company operates. I'm sure other momentum tech stocks were subject to scrutiny and market winds all the while trying to change the paradigm of your business or create a new one...
racer26
Active Member
Doubt it - the hardware change is a few extra cameras, amounting to maybe $300-500 per car. The price tag of AP went up by $2500, so if anything it should be accretive to profit.
jhm
Well-Known Member
That's what I'm talking about. A third advantage here is that placing the peaker within the substation means that the substation can draw less power from the transmission grid at times of highest demand. Thus, the peak load on the transmission system is reduced. So there is a triple win here:
Generation -- reduced need for centralized peak power generation
Transmission -- reduced load on transmission grid at peak times
Distribution -- reduced peak load on transformers or expanded power capacity without transform upgrades
This means most of the infrastructure of the grid can be better utilized. Generation, transmission, and transformers can move closer to constant load, which increases there capacity factors. For example, CAISO runs about 25 GW average over the year but close to 50 GW peak. As demand for power in California grows (say due to massive EV sales!) that average load could approach 35 GW. What happens to the peak? If it scales up to 70 GW, then the state will have to make massive investments in expanding the transmission grid. But if lots of storage happens at the substation level and closer to consumption, then perhaps the peak stays near 50 GW. It's capacity factor rises from 50% to 70%, which makes the transmission grid much more cost effective. The counterpoint to this is that centralized generation must also become much more base load like in aggregate, including all that solar and wind. So I could see some role for pairing central power production with storage so that it can be transmitted at a more uniform rate. This is in part why I am skeptical of proposals to use a massive HVDC backbone to balance intermittent renewables from different regions. To use transmission lines this way would seem to lead to low capacity factors for the lines. They basically sit idle waiting for the wind to blow harder in one region than another. However, if you buffer renewables with storage before transmission, then such a backbone could have near constant load and very high utilization. The paradox here is that if you need to use batteries to buffer both before and after transmission so as to optimize throughput, might you be better off just using storage to avoid transmission? To be sure there are economic trades that would suggest some balance of transmission and storage pre and post transmission is optimal, but all this is looking out to a time when CAISO will have over 50 GW of batteries to play with. For today, I am just happy to see that PG&E is exploring the wide benefits of substation storage and believe that is the best place to start accumulating some 25 GW of batteries plus EV batteries.
mmd
Active Member
Few weeks back, we were discussing that order backlog should be good as the delivery date was many weeks away. Today it's Feb 17. Next 7 (or 10?) days, there is a factory shutdown. The delivery estimate here in CA is still 'Late March'. Feb is a 28 day month.
What can we infer from this?
What can we infer from this?
I doubt very much that Elon's opinion has changed and probably has become even stronger given how volatile/crazy the markets have been with SCTY and TSLA. But when I first read your post I thought the article was quoting a recent email so thought it was worth clarifying so others would not also be confused.
It's a couple extra cameras and replacing the Mobileye chip with the much more expensive nVidia PX2 unit, which, I'm quite sure, costs more than the extra cameras themselves.
MitchJi
Trying to learn kindness, patience & forgiveness
They are trying to figure that out:
Tesla plugs big batteries into PG&E’s electric grid
PG&E and Tesla, for example, are also collaborating on a pilot project that uses Powerwall batteries in homes and businesses in the Bay Area. The amount of energy those scattered batteries can store will be equal to the amount stored at the Browns Valley substation, Della Penna said. PG&E will be able to study how both setups — one centralized, the other diffuse — affect the grid.
“They’re each with their own challenges and opportunities,” he said. “We’ve structured it so we’ll have a lot of really good learning here.”
Last edited:
racer26
Active Member
Do you have any evidence to support that the custom solution (derived from the Titan GPU that DrivePX2 is based on) is significantly more expensive than the EyeQ2 that was in the MBLY based solution?
We beat this to death a couple of days ago - the NVDA hardware that AP2 runs on is not costing more than $1000/car. Its based on a GPU whose reference board in full retail trim for PC gaming use in quantity 1 is retailing for $1000. In customized trim for mass-production, its likely costing half that.
Is the overall AP2 NVDA based solution more expensive than the AP1 MBLY based solution? Yes. It probably is - it has 7 more cameras and a more powerful chip. Overall, the whole hardware suite is likely about $800-1200 more expensive than the MBLY-based solution. But, as I pointed out - the convenience features it enables are priced $2500 higher than they used to be - AP2's EAP option is $5,000. AP1's price in 3Q16 was $2500. Both have near 100% take rates.
The differential cost of AP2 will not significantly impact the bottom line in a negative way, as it is impacting revenues as well. If anything, AP2 will be a net positive to the bottom line.
racer26
Active Member
Its only late March for Model S in USA. US Model X is showing May.
I have not challenged your assertion that production capacity for Model S is now approximately equal to demand. The thing is that because Model S and X share a final assembly line, the total demand for S/X is the important factor. The backlog is not necessarily a good thing - it means people have to wait to get a car. Some fraction of buyers aren't willing or are unable to do that. They want or need a car now, and they'll go to Mercedes or BMW or someone else if Tesla cannot provide.
Backlog is good in that it means you definitely have buyers for every car rolling off the line. Its bad in that it means you're turning people away. X being backlogged and S being relatively more accessible is kind of a best of both worlds type scenario. Want a car as soon as possible? Buy an S. If not enough people buy an S, then we just compensate by burning through the X backlog a bit faster.
DurandalAI
Member
I don't think 3 months out for getting a Model X is extraordinary at this point. Ordering a car from the factory is going to be a 1-3 month wait from many car manufacturers. And there are a number of New Inventory units available for those impatient buyers.
Buy a new Tesla today | Tesla
Edit:
Added a link about special ordering a Mercedes.. Typical wait time being 6-10 weeks
Should you special order your new Mercedes-Benz? | Cartelligent
Last edited:
I don't think 3 months out for getting a Model X is extraordinary at this point. Ordering a car from the factory is going to be a 1-3 month wait from many car manufacturers. And there are a number of New Inventory units available for those impatient buyers.
Buy a new Tesla today | Tesla
Edit:
Added a link about special ordering a Mercedes.. Typical wait time being 6-10 weeks
Should you special order your new Mercedes-Benz? | Cartelligent
This is not an equivalent comparison. For many other automakers, they have enough volume capacity to switch to building other models. The model you ordered may not even be scheduled to be built for a month at that plant. Further, your order goes in amongst other vehicles ordered by dealers. The Tesla factory builds mostly custom spec cars. The total amount of inventory cars is still a small fraction of the overall build numbers while it is flipped for Mercedes. You can't compare in this way.
If the heart of the issue is demand, you have to factor in Tesla's wait queue differently because the percentage of people that buy off the lot versus custom order is very different. The amount of inventory sitting on lots therefore is completely different magnitude and the resulting analysis of demand has to be adjusted.
Right now, there are about 1,000 Model S/X listed as demos globally. There is some more inventory that are showroom vehicles that not necessarily listed for immediate sale. Similarly, some demos are intentionally reserved for various reasons. There are about 265 retail and service locations. Assuming that 200 of them are retail or retail and service, and 65 are dedicated service, that's 5 vehicles per location. That's not a lot of vehicles. In comparison, there are 12,814 Mercedes E class vehicles listed by dealerships on Autotrader right now in the U.S. That's for a vehicle with 51,000 in sales in the U.S. last year. So basically, 3 months worth of sales is sitting on the lots right now. If that was true of the Model S, that would be about 7,000 Model S inventory vehicles sitting around U.S. Tesla retail locations right now, using InsideEV's estimate of U.S. sales. Clearly not the same thing.
Last edited:
jhm
Well-Known Member
That's super. Within a substation, we can make a distinction between centralized batteries and behind-the-meter batteries in homes and businesses. I suspect that both will be quite valuable. Behind-the-meter has added benefits for distribution such as voltage regulation which can be worth close to 1c/kWh, integration with rooftop solar and greater reliability for the consumer. But you do need to aggregate these and have complexity about who controls what, but that needs to be worked out anyway to create good adaptive EV charging too. So I suspect that utilities will want to support both behind-the-meter and front-of-the-meter storage at the substation level.
MartinAustin
Active Member
I believe Elon tweeted last year around April that they were running with an ASP of $42,000 for the Model 3.
Thanks, I am aware of this.
I, however, was using assumption for the 70kWh battery pack, so bumped up the ASP - to account that this ASP is not for the whole model lineup, which most likely will be heavily tilted toward 55kWh pack, but specifically for 70kwh variants.
RobStark
Well-Known Member
I believe that is the expected median price.
Many mucho loco Teslanarians will be buying fully loaded performance Model 3 raising the average selling price.
- Status
Similar threads
- Sticky
- Locked
