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

2017 Investor Roundtable:General Discussion

This site may earn commission on affiliate links.
Status
Not open for further replies.
So at this point, Section A assembles cells into packs using both Gigafactory and outside cells, sections B and C produce cells, and D, E, D' and E' are to support cell and pack assembly for 35/50 GWh, and F is for automotive subassembly (Model 3 drive units).

So... with Sections D, E, D' and E' looking like they are ready for an "inside" then Tesla is almost done with the capital expenditures necessary to achieve the original 35 GWh of cell production. It is Panasonic that needs to spend the rest of the $1.6 billion and other partners for the cell production and Tesla does needs to spend money on the pack production that point. So Tesla has spent $1.98 billion already... but that also includes Section F, which wasn't in the original plans.

Given about 9 months of installation and commissioning, I am now expecting that Tesla is looking for 35 GWh of cell production in mid 2018, not end of 2018.

The entire 2020 plan which includes 35 GWh of cell output and $2 billion invested now looks 2 to 2.5 years early and with only 30% of the building footprint built out. Again, 100% of the cell production capacity but with 30% of the building footprint and that includes a section doing something that was not part of the original plans.

Now, it is reasonable to extrapolate that the 5k/week build plan for the Model 3 has to be satisfied by the pilot phase (Sections A, B, and C). That means roughly 16-17 GWh of cell output is required out of the pilot phase. If Tesla is sourcing Samsung cells for a project in Q3/Q4, that means they are looking to achieve that output earlier rather than later.

Also... the size of D, E, D', and E' is bigger than A, B, and C. That's notable. We don't know what that means yet, but I suspect 35 GWh cell output is not the actual expected end capacity and the expected output is closer to 40 GWh.

First off, thanks so much for putting together this comprehensive post. It is extremely helpful.

At the time of the Gigafactory Party one year ago, the exterior walls were [mostly] up for Section D but the floor was open rebar awaiting the concrete. D', E and E' didn't exist - there were just the beginnings of the footings for those sections. Given the timing I'm wondering if Tesla needs Section D producing cells (in addition to B and C) to reach 5000 M3 per week and is depending on D', E and E', which lag by 6-9 months to reach 10,000 per week in 2018.
 
  • Like
Reactions: jbcarioca
I have discussed Tesla with one Apple software engineer who moved over there. He was more horrified than impressed, but I'm hoping his experience was not typical. My impression is that there's lots of cowboy code hacking and very little software engineering. That works for a while, but it's not sustainable. We shall see, but I'm a bit worried. I am not at all impressed with the progress in the auto-pilot software since AP2 was released.

It is somewhat counter intuitive, but that is exactly what you want in a more startup type environment. Much like going into debt every quarter is not sustainable you want to create lots of technical debt in your startup phase.
 
h
From my many years doing software development at Apple (not currently), I'd say they rate as somewhat decent on software engineering. Without getting into details, they are another Silicon Valley company where shipping dates reign supreme and software quality is only slightly relevant.
Recognizing and hiring great people, sure. But I don't think Elon has much understanding of large scale software development. Remember, in software there are no "first principles" so his favorite mental tool doesn't work.
Giving examples of what SpaceX has accomplished says little or nothing about Tesla. Pushing updates to Tesla vehicles is essentially the same thing Apple has been doing for years pushing software to its devices, so is not evidence of greater capability. I have no idea what you mean by "full spectrum".
I am not at all impressed with the progress in the auto-pilot software since AP2 was released.

By 'full spectrum' software development I mean there are many different tasks SW is built to accomplish and the exact skills and best practices to be successful in different spheres are not identical. Rigorous adherence to the tenants of software engineering are vital in some of those but are not closely followed in others due to cost, time and business constraints. Don't assume Musk hasn't much understanding of large scale software development. He doubtless has well founded views on how and when SE best practices can be short changed in order for the more critical business objectives to succeed. Since Tesla and SpaceX were founded and grown by the same man who made his early fortune building software companies, and SW is mission vital to both, it's logical to take SW technical successes by one as demonstrating the other companies successes are due to competence not luck. There have been many posts explaining why AP2 has been slow to replicate AP1 features and why AP2 progress has likely accelerated. That there have been poorly executed Tesla SW based features is a fact, but if those were the rule rather than exceptions Tesla would not have evolved successful and compelling features for their cars. Musk is more interested in pushing the envelope to hasten change than he is with achieving a 5 in the capability maturity model. In this he has his priorities in order.
 
Smart brand’s move to all-electric results in losing 2/3 of US dealerships

"<
Automotive News reported that out of the 85 Mercedes-Benz dealership in the US that were carrying Smart cars, only 27 decided to keep selling them:

“Mercedes-Benz USA, which distributes Smart in the U.S., asked dealers to decide by the end of June whether to continue selling the ForTwo two-seater, the brand’s sole nameplate. Of Smart’s 85 outlets, 27 said they would remain, while 58 said they would move to a service-only operation, said company spokeswoman Donna Boland. Those numbers are preliminary, she said.”

The remaining dealerships are mainly located in California, New York, and other states with zero-emission vehicle mandates.
>"

like I said-
ICE-Co doing the best they can - but they're in the wrong business and the retail stores know it
the grind of disruption continues apace
 
Smart brand’s move to all-electric results in losing 2/3 of US dealerships

"<
Automotive News reported that out of the 85 Mercedes-Benz dealership in the US that were carrying Smart cars, only 27 decided to keep selling them:

“Mercedes-Benz USA, which distributes Smart in the U.S., asked dealers to decide by the end of June whether to continue selling the ForTwo two-seater, the brand’s sole nameplate. Of Smart’s 85 outlets, 27 said they would remain, while 58 said they would move to a service-only operation, said company spokeswoman Donna Boland. Those numbers are preliminary, she said.”

The remaining dealerships are mainly located in California, New York, and other states with zero-emission vehicle mandates.
>"

like I said-
ICE-Co doing the best they can - but they're in the wrong business and the retail stores know it
the grind of disruption continues apace

Maybe they could contract out with Tesla to sell their EVs...
 
  • Funny
Reactions: Skryll
Really? So you're the guy buying a 2012 Model S because it will keep you happy until 2027?

Certainly sounds reasonable to me. My '91 CRX has kept my wife and I happy through 2013, and we still use it today. This year's Model X will hopefully be the car I'm driving 40 years from now when they take away the fob from me :) Maybe I only get 20 years though ...
 
From my many years doing software development at Apple (not currently), I'd say they rate as somewhat decent on software engineering. Without getting into details, they are another Silicon Valley company where shipping dates reign supreme and software quality is only slightly relevant.


Recognizing and hiring great people, sure. But I don't think Elon has much understanding of large scale software development. Remember, in software there are no "first principles" so his favorite mental tool doesn't work.


Giving examples of what SpaceX has accomplished says little or nothing about Tesla. Pushing updates to Tesla vehicles is essentially the same thing Apple has been doing for years pushing software to its devices, so is not evidence of greater capability. I have no idea what you mean by "full spectrum".

For Tesla what measures their software engineering prowess is whether they can update their existing software into something better. That stresses their source code control, testing, and release processes. So far the evidence is not very good. From the outside it looks like much of their code is sufficiently messy that their regression rate is high -- as they fix bugs they introduce as many new bugs, and as they add features they break existing features. Some things are so ugly (e.g. the music UI) that they don't even try to fix it -- that's what you see when the guy who wrote the code is gone and nobody dares touch it.

I have discussed Tesla with one Apple software engineer who moved over there. He was more horrified than impressed, but I'm hoping his experience was not typical. My impression is that there's lots of cowboy code hacking and very little software engineering. That works for a while, but it's not sustainable. We shall see, but I'm a bit worried. I am not at all impressed with the progress in the auto-pilot software since AP2 was released.

If you look at the new navigation interface, it is starting to show some abstraction that should make it more easily maintained.

I think the swift guy helped.
 
  • Helpful
Reactions: Ulmo
Many auto manufacturers are likely looking at this more and more-- current shopping paradigm is going online for many purchases, and auto manufacturers could see an opportunity to recover extra margin...
The dilemma for other manufacturers is without dealership, they will also loose their vast n/w of Service centers. That was how things got set up the old way
 
  • Helpful
Reactions: ValueAnalyst
The dilemma for other manufacturers is without dealership, they will also loose their vast n/w of Service centers. That was how things got set up the old way

that would be an interesting lawsuit by dealers in various states; not sure those laws allow manufacturers to go direct if ANY dealer in state is selling- State dependent of course- but....

So ultimately, even if legacy automakers go whole hog into EVs, they will be constrained by the dealership network (at least in the US)... who will refuse to sell them. Unless the automakers completely STOP selling and making ICE vehicles.

Another Tesla moat.
 
Possibly traditional auto manufacturers could see service centers as a profit generating source, and would quickly cut out traditional dealers/service centers.

Thats until Tesla comes along(already did) and says goal of service center is not to make revenue ;)

Also dealerships will not let go without a fight, unless the manufacturer is badly in the red ..
 
  • Like
Reactions: Lessmog and larmor
Possibly traditional auto manufacturers could see service centers as a profit generating source, and would quickly cut out traditional dealers/service centers.
Service, yes perhaps. But there seem to be fiercely defended legislation prohibiting manufactures selling direct, certainly if they already have any dealers, in most if not all of the US. They will probably not lose the service revenue willingly either.
 
Certainly sounds reasonable to me. My '91 CRX has kept my wife and I happy through 2013, and we still use it today. This year's Model X will hopefully be the car I'm driving 40 years from now when they take away the fob from me :) Maybe I only get 20 years though ...
And I was happy with my '94 Accord until 2013 when I gave it to a relative. But that's irrelevant. These new things aren't cars, they're computers on wheels. They will obsolete much, much faster.
 
Status
Not open for further replies.