Some quick thoughts about the truck Elon talked about in the Recode podcast:
1. I don’t see where Tesla would get the batteries for a high-volume truck in the next few years. 3 & Y should soak up most of GF1’s production, with Semi, Roadster, and TE using any spare capacity. As such, Tesla’s first Pickup almost has to be high-price, high margin, low volume.
2. As a low volume project, it can’t command a lot of CapEx. It should be able to share batteries & motors with 3/Y/Semi. It also needs to share production space— either with the S/X production line, or with the Semi production line.
3. I have no idea if it’s technically feasible, but I’d love to see it made on the S/X line. It looks like the line can produce 30k vehicles per quarter (120k/year) but Tesla is 18650 battery limited to ~100k per year. If Tesla could produce 20k/year Crazy CyberPunk Trucks (using 2170 batteries) on the S/X line, they’d more fully utilize their production capacity & increase margins on all three vehicles.
Well, I don't think they are going to build Y in Fremont, at least not only in Fremont. They're not going to build 10k/week+ in Fremont. I think they need another NA GF for Y+Pickup at minimum, possibly two if they can't do some production at Fremont. Those GF's will get their own cell lines with enough capacity to feed them. GF1 only has to feed Fremont (3, S/X after switch to 2170, anything else that gets built there) and GF1 (TE, anything else that gets built there).
Building the pickup on S/X line is a bad choice, at least without a major overhaul to the S/X platform (which arguably they're due for anyways), and even then I don't think there's really a good argument for sharing S/X/pickup. Just because other makes share SUV and truck lines doesn't mean you can easily build a truck on the S/X platform. Certainly if they did share a platform, S/X should switch to 2170 before or at the same time, not be running a mix - and then your argument to make a paltry tends of thousands of high priced trucks to take up slack in the production capacity (due to cell supply) would be solved by just building more S/X anyways.
What I'd like to see is that GF3 (China) and GF4 (Europe) move ahead on their own schedules, financed through local debt (like China), first doing GA of Model 3 and possibly Tilburg style assembly for S/X, and planning ahead to have the cell capacity lined up not only to bring up 3 production but also Y production at each location.
Separately, in parallel, I'd like to see construction for GF5 and possibly GF6 (if Fremont/GF1 can't handle the added production) somewhere in North America, someplace where housing workers isn't such a problem - perhaps targeting someplace where some existing manufacturers will be failing so that it's easy to pick up labor. GF5/6 should be georaphically spread apart - perhaps one in the rust belt, the other in the south. These will produce Y and pickup to start with, perhaps Semi. These may have to be built from cash (rather than debt) unlike GF3/4, so likely have to start one at a time. I would get Panasonic to install 1-2 lines ASAP and start with making TE cells and growing TE output. Once almost ready to start Y and/or pickup production, switch one or more cell lines to vehicle chemistry for a short while (perhaps a week or two?), then switch back to TE. This brief output will be enough to do initial start of ramping, and once you get close to needing a whole cell line's output, switch it over to vehicle output again, and have Panasonic install another pair of lines. The new lines may not even go to TE initially, depending on timing, may already have switched to pure vehicle output from the cell lines and need more.
Ideally, Panasonic cell line installation should stay ahead of the product curve and TE should consume the excess, which ideally should cover the costs (even if margins aren't great compared to vehicles) of the Panasonic lines, so that less cash is needed overall. As production ramps, switch cell lines as needed, and keep adding more cell lines. Eventually TE production will be plentiful once lines are ramped at all GF's as the total cell production capacity should be well beyond vehicle needs.
TL;DR : Pickup and Y should be high volume, and probably need at least two North American production locations (GF5 and either GF6 or what's left of Fremont+GF1), necessitating at least one if not two new NA GF's to be built. I think that by intelligently building cell lines ahead of production and using them for TE that a good portion of the costs can be covered, reducing the need for cash from Model 3/S/X sales. GF3 (China) / GF4 (Europe) should be largely funded by local debt and thus not impact this (but I assume Tesla still can't get favorable debt financing because Wall St and must use cash for NA growth). S/X line under-utilization will be fixed when they move to 2170 cells from GF1, and S/X are not good truck platforms. Semi may be built at any of the above locations, Roadster likely only at Fremont. At the rate of cashflow generation they're getting from Model 3 I don't think any of these are a problem in terms of costs, only in terms of timelines (time to build the building that houses the machine to build the machine ... ) if they want to get to volume production of any of Y/pickup/Semi in 2020. If they can get the buildings built in the next 6-8 months I don't see why they can't at least be doing initial start of production at the 10's or 100's per week range of Y/pickup/Semi by end of 2019, assuming they've been planning ahead and really learned from Model 3 this time (like they claimed to have learned from X, and S before it...).