Single data point, but that is why I'm on my third Model S and not still driving the first one.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
- Status
Single data point, but that is why I'm on my third Model S and not still driving the first one.
Krugerrand
Meow
Meh. It's possible to have that affect, but people have really short memories. And if you can dazzle them, they have no memory at all. In a perfect world we want unbelievable results AND perfection in those results. Never happens.
I am very confident that S/X sales will increase 50% again this year. Besides new markets, Tesla has a draw that no other car company has - rapidly evolving tech. I have always kept my cars a minimum of 5 years. But with the Model S, I have upgraded every 2, and am wringing my hands to not upgrade again! It is worse than a new computer or smart phone. If you love gadgets, which I'm sure a lot of Tesla owners do, then Tesla will be selling a lot of cars to repeat buyers every 1-3 years who just can't resist the rapid evolution of Tesla vehicles. This is yet another game changer that most people haven't factored in to the demand picture yet. But I have learned my lesson, and for the first time in my life, I think I will lease next time....
Krugerrand
Meow
Let me try and recall from my memory why there is such a long lead time for this and why it's not like slapping a bunch of Legos together.
First you design the car and break down its individual parts.
Then you have to design the die sets to make each part (yes, SETS as in each part will require 4-5 individual dies to make the part).
There are die standards that have to be followed and are extensive.
There is a slew of information that is required that the design has to fall within like; what is the exact material (type, thickness, supplier) that will be used, what pressline will be used and what are those specifics (size, tonnage, stroke rate, shut height, cushion pressure range, clearance for robots, scrap chutes, stackers etc...).
Once all of that is taken into account then you have to run computer simulations. In the old days there was no such thing as computer simulations, so there is some chance to cut out valuable time here and down the road via mitigation of possible problems. However, there is such a thing as bad die design, in which case no amount of changing the die design will fix it and you then have to go back to the drawing board.
During this whole process there will be die design reviews FOR EVERY SINGLE DIE (we can be talking upwards of 200), checking the progress of the design, making sure something hasn't been overlooked, something isn't missing, a mistake hasn't been made, the die standards are being followed etc...
The next step is then to order the castings. I can't remember the average turn around time for castings, but it's got to be a few weeks from start to finish for one die - including shipping time.
Then you get to start making the blocks for the die. In the old days that was all done by hand. Machines and computers makes that process a bit faster, but not a lot because you still need someone to program the machines - and if they get it wrong, you get to start all over again.
So you make all the blocks, punches and assorted bits and pieces and 'build the die'.
During this time there will be die reviews making sure die standards are being followed, that progress is on time, blah, blah, blah.
Then you get to take it apart and send all the pieces out to be heat treated. That mostly goes okay, but sometimes it doesn't. Sometimes a block will split, get damaged, get 'misplaced' and you get to make it all over again.
You get all your blocks and parts back and you get to put the die back together again. Then you get to put it in a try out press and see if it works. I can tell you, you never get a part in spec at this stage because the computer simulation is never 100%, because mistakes get made, because sometimes the material supplier goes out of business or no longer makes that material or changes their formula, or any number of other things that can and often do go wrong.
So at this stage you get to make changes to the die. Sometimes it's a change to a cutting edge, sometimes it's a different punch, sometimes it's grinding down a block or welding a block up and reshaping it. Sometimes it's making slight changes to the press, increasing or decreasing cushion pressure, slowing down the stroke rate, increasing or decreasing tonnage, blah, blah, blah. Hopefully it's a quick and easy fix - lots of times it's not. If the part is simple, sweet. If the part is complicated, not sweet.
And then when you've produced a part that's 'mostly' within the company's specs (there's a percentage range that companies will accept, (ie. might be 80% of the marked points have to be in spec) they show up to do yet another die review. If all the criteria is met, you ship it (takes time) to them and then they get to go through a whole set of homeline trials (typically 3 stages), which can take weeks to get through to produce a part good enough to make a car with. Why weeks? Because the dies were most assuredly NOT tested on the exact same press, using the exact same material (as in from the same supplier) as when they end up on at the company. So you don't just get to slap them in and voila they work. Tesla may be able to mitigate some of this because they own the tool and die company, but I can guarantee that one company is not big enough to be making all the Model 3 die sets - certainly not in the time frame.
I've probably missed some steps in there, but that should give you the gist. It's a long, tedious, complicated process. And there's a whole other level of exactness required to produce a die set that makes a Class A body panel, than a die set that makes TVs or barbecues.
Last edited:
What we do know about guidance for 2017 vs. guidance made for 2016 a year ago:
net S/X orders in the most recent quarter is 52% more as Tesla sets guidance imminently vs. when they set it at 80-90K units last year
What we do not know:
1) how much of last quarter's record demand was due to the ending of free SuperCharging for life
2) how much will the release of the Model 3 on the horizon impact S/X demand
3) what improvements to the S/X may be coming this year to spur demand
my guess is they deliver 110K S/X this year with demand being a bit higher than that. perhaps it would be interesting to have a thread for predictions on total revenue for 2017, in addition to the existing thread on predicting deliveries. I think it's worth noting that as of today the average analyst revenue forecast for 2017 is $9.89 Billion per Yahoo Finance.
Lease... then you are locked in to maybe 3 years... what happens when something you *MUST* have comes out and you are only 1 year in to the lease?
Just accept who you are and get on with it I say! Besides, any money you plan to leave to your children they will likely just squander anyway...
Mike
Seems like a good time to repost this article from September on a supplier of Model 3 dies:
The company that helps Tesla make aluminum look sexy- Nikkei Asian Review
I like the way the author used past tense for the creation of Model 3 dies. Of course, there are many dies to make, but still...
The company that helps Tesla make aluminum look sexy- Nikkei Asian Review
I like the way the author used past tense for the creation of Model 3 dies. Of course, there are many dies to make, but still...
I've written about this a number of times. They are morons. Revenue of $9.89 Billion means no increase in S/X deliveries, no Model 3, and no Tesla Energy. I'm predicting that revenue will be $15-20 Billion this year, and stock hits 300.
If I remember correctly, if you're planning to have multiple production lines, you actually want to make the inverse castings for *making* the blocks and other components of the dies (so that you can make multiple copies of the die fairly easily). These are made by very exact machining etc., and take a long time to make. Then you can actually manufacture the dies rather than making them as one-offs...
Dunno if Elon's doing this, but with the "factory as a product" attitude, probably.
Anyway, this adds a whole 'nother layer to the lead time.
hadn't noticed your earlier comments, but, yes, worth noting, this will be a year where it's all but certain those numbers have to go up very substantially, repeatedly. the way these analysts play, case by case such revisions may or may not lead to higher price targets for any particular analyst... but, for those with "buys" on Tesla, if the stock price passes their target, that passing will in nearly all cases lead to maintaining a buy and raising their price targets.
Krugerrand
Meow
Great article that touches on some of the complicated aspects that simply take time to do.
N5329K
Active Member
I wonder how many computer users who required the services of Dell Support ever bought another Dell?
Robin
Krugerrand
Meow
I'm not sure I'm understanding what you are saying. Are you suggesting that Tesla might run multiple press lines of the same part at the same time and therefore need two sets of the dies?
Adding: Or are you suggesting making multiples of the same blocks for dies within the same set?
Krugerrand
Meow
CALGARYARSENAL
Member
After a tour of gigafactory, Whitney Tilson decides Elon and J.B. are the real deal and would never bet against them again.
Whitney Tilson Was THIS Close To Shorting Tesla .... Again - ValueWalk
Whitney Tilson Was THIS Close To Shorting Tesla .... Again - ValueWalk
Thank you for pointing that out. It is really a remarkably low number -- only a 9% increase over Q3 2016. I would expect Tesla to be able to hit that number on increased S/X sales alone (my 2017 estimate is similar to yours), with no contribution from Model 3 or TE ramp.
In fairness, Tesla is playing TE close to the vest and I expect many analysts will wait for concrete guidance before they will factor significant TE numbers into their forecasts.
Last edited:
Yes.
*As a special case*, if Tesla plans to open a European factory, and a second American factory, and an Asian factory... they might need at least 4 sets of every die -- one per factory. If there's a part where there are a large number used in each car, they might need to run it on multiple press lines at the same time just to keep up (though I'm not sure if there are any of those). And dies wear out after a while and they'll need replacement dies every few years (possibly sooner).
Awesome find!
Note: hidden within is the 30-minute Q/A with Musk!
I recommend downloading it in case it disappears:
http://www.valuewalk.com/wp-content/uploads/2017/01/TeslaQA.mp3
Last edited:
I want to second this as a very good read! Thanks @CALGARYARSENAL
Another (of many) good quote from Tilson:
Two top engineering guys I met a couple of years ago who work/worked at Telsa told me that JB Straubel is also the real deal. One said to me: “There is NOTHING that Musk and Straubel can’t do”. Why would I want to be short BOTH of them?
This is a terrific article. Nothing really new for long-term bulls, but a very nice summary of all the reasons not to bet against Tesla from someone who I think by his nature tends to see things from a bearish perspective.
Loved the summary, maybe because it reminded me my own sense of deja vu -- this really does feel like 2012 (not quite early 2013 yet) all over again:
In summary, today reminds me of early 2013, when the stock was at $35: the company was/is burning a ton of cash, has missed all sorts of deadlines, and there was/is much skepticism about a major new product launch (then: Model S; now: Model 3). Back then, I figured Musk/Tesla would have to pull an inside straight to beat me – but they did. In hindsight, I don’t think it was luck. They were much better engineers than I thought and built a brilliant, innovative car. Tesla is chock full today of even more, better engineers, led by the same two guys, so why would I bet against them now? Fool me once, shame on you. Fool me twice, shame on me
Last edited:
- Status
Similar threads
- Sticky
- Locked
