TSLA Market Action: 2018 Investor Roundtable

[Matias]

It's not going to be cashflow negative for that long, only the first year or so until they get , especially from wealthy customers in the finance industry where a few less milliseconds of lag is important.

Why would it be faster to beam information up to LEO and down than use optic fiber or microwave link?

Edit: microwave is faster than optic fiber, because electromagnetic radiation moves faster in air than in fiber, but I don't see advange satellite vs microwave link. Any HFT is a niche application.

 

[hobbes]

I guess he did mean by next year, I thought he meant that it would be available sometime next year: Elon Musk on Twitter

Things must be moving faster than we thought... (Or at least than I thought.)

I think some people are getting confused over his tweets - here he is clearly talking about two different releases, one soon and one sometime next year:

For those unfamiliar, this uses Tesla Autopark/Summon. Slightly smarter version hopefully ready soon. By next year, a Tesla should be able to drive around a parking lot, find an empty spot, read signs to confirm it’s valid & park.

And the most recent tweets (RC/follow like a dog) refer to the first timeline, so the car won´t be driving around looking for a parking spot in 6 weeks. What I found misleading in the latest tweets was that the car supposedly will drive to your phone. That could be anything from coast-to-coast to from garage-to-driveway (what it does already). I think this will be very limited (as in: not on public roads/in traffic) in the beginning and the new thing is really that it is based on location of your phone, not your manual input.

 

[ItsNotAboutTheMoney]

I think one of the primary goals of Starlink is as a dry run for Mars. Seems much of Tesla/SpaceX/Boring Co tech is developmental work for Mars. Starlink will be great in a planet without large oceans where for a modest expense a global martian comm link plus GPS can be established. I do think that some of the Starlink tech can be used in the vehicles but perhaps more in conjunction with terrestrial or inter-vehicle linking. I wonder if starlink could be used as a private GPS system at greater precision?

While I think it's very useful for Mars, I think Starlink and other similar projects are simply the result of easier satellite downsizing, lower satellite costs and lower launch costs. Technological advances combined with cost reduction brings disruption.

Whichever companies get there, there's going to be radical change. I look forward to the end of rural broadband subsidies, and the positive impact of cheaper, better communications for island communities. Combine that with cheaper solar and batteries and I think islands are going to be more popular. Unfortunately, with that will come even more cases of local young people being priced out of the market by outsiders buying vacation homes.

 

[Ellec]

Can the starlink posters post in the dedicated thread please, this is Tesla Market Action.

 

[X Fan]

Premarket was back up to just under $340 and autos doing well so far with bmw & Toyota up.

Looking forward to a positive day....

Edit: corrected $ amount.

 

[LST]

Lots of technologies don't work 'somewhere', but work great 'elsewhere'. Bicycles are great for Canberra, but not Northern Sydney (too hilly). Evaporative a/c works in the dry west, but not the humid coast. We can't write it off because some places aren't suitable for version 1.

Sure that's correct, but speaking of Europe as "some places" thus implying it's not an important market is a little bit of stretch, we are talking of a place of ~750 million inhabitants...And i guess my concerns are valid in urban areas all over the world. Not that i think Tesla should not give it a try... in some places suitable.

 

[Tslynk67]

Premarket was back up to just under $350 and autos doing well so far with bmw & Toyota up.

Looking forward to a positive day....

$350? Did you mean 340??

 

[avoigt]

Sure that's correct, but speaking of Europe as "some places" thus implying it's not an important market is a little bit of stretch, we are talking of a place of ~750 million inhabitants...And i guess my concerns are valid in urban areas all over the world. Not that i think Tesla should not give it a try... in some places suitable.

Its my understanding that some places refers to some places in Europe and not Europe to be some places.

Actually I agree to that view. Every step AP makes will lead to situations that in some places it works better than in others and it may not even to be able to be used at some others until it improves. Its not the idea to make it work at all places at once.

This is an interim status and you just have to wait and see until autonomous driving will catch up. That sentence applies for all systems with neural nets and enough data to operate beside some other restrictions hence it will not be applicable for all companies offering an autonomous system.

Its a funny concept of human kind to pick the few situations a system does not work and call it the system does not work. Well thats true but only in that specific situation and at that specific time. If it works in most of other situation it still does add value to most.

I realize you live in Switzerland and understand where you comment about narrow roads is coming from. There are many old city towns that are not really designed for cars but you certainly will agree that the vast majority of all roads in Switzerland are on a very different level where the system will operate just fine.

 

[winfield100]

!!! a hyperloop toy !!!
electric car that goes _really_ fast
Zipster™ Gravity-Defying Speed Pipes

Zipster™ Gravity-Defying Speed Pipes

 

[X Fan]

$350? Did you mean 340??

View attachment 348869

Oops...fingers had not woken up yet (nor my eyes I see).

Thx for correction.

 

[neroden]

• I believe the 'driving' logic will simply detect that scenario in your town and declare the road a dead end, stop the car and alert the driver - or if there's no driver alert a remote driver or route around it using another approach.

Fine, but that's not level 5 "full self driving". That's level 3 "partial self driving". A nice driver assist feature. But with this, *you need a driver in the car at all times*, and "autonomous taxis" to ferry drunks around without a driver are impossible. My point is made.

P.S. I should repeat that I think limited self-driving -- either geofenced to particular expressways, or bailing out in bad situations -- is actually relatively close. This isn't the dream of "sleep while you drive" which so many people want, however.

 

[neroden]

Heads have rolled at SpaceX.

Musk shakes up SpaceX in race to make satellite launch window: sources | Reuters

How does this affect the Tesla Network?

It doesn't. In this case, I think Musk is right; the whole point of these satellites is that they're cheap (i.e. if they break, put up more); the test-launched satellites work, and going through extra rounds of testing is simply wasting time. What this will affect is Comcast, Verizon, et al. I'll be buying that "pizza box" satellite connection as soon as I can. It'll have a waiting list longer than the solar roof, possibly longer than Model 3.

 

[avoigt]

Step by step and one after the other ....

Upgrading Tesla To A Buy, $425 PT On Profitability

I have been privately and publicly bearish on Tesla's stock. But after the phenomenal Q3 report, I believe the profitability and cash flows generated are now sustainable.

Upgrading Tesla from a sell to a buy while putting a $425 price target on the stock.

This price-target is based on a DCF model that excludes any financial impact from inevitable future product innovations like the Model Y, Tesla Semi, and Roadster.

The goalposts of the bear argument have now moved from profitability/cash generation to the valuation on the profits, putting the massive bearish community on Tesla in a precarious position.

If Tesla is able to show sustainable profits and cash flow like I believe they will, then many short sellers will most likely cover, providing support for the stock.


Its a SA article and I decided not to link or promote that page for obvious reasons therefore no link.

 

[Fact Checking]

Why would it be faster to beam information up to LEO and down than use optic fiber or microwave link?

Because through space and the atmosphere the radio waves and the laser beams Starlink uses are propagating at over 99.9% of the speed of light, and in a straight line between the satellites, through very few 'hops'.

Fiber optic cables on the other hand only run at about 68% speed of light, and they also have a lot more 'hops' typically to boost the signal but also to route around physical geographical features.

Note that for example New York to Frankfurt ping times over fiber optics are 96 milliseconds: that's 28,800 km at the speed of light round-trip. The two cities are only 12,400 km round-trip apart when taking the curvature of Earth into account - which means that even in practice fiber optical networks between two of the largest financial centers of the world are very sub-optimal. The distance of bouncing up 400-600 km to LEO distance is a minor cost of latency: the path of photons will still be well below 20,000 km and the LEO communications latency will beat the fiber optical latency.

If we go further down the hierarchy of geolocational advantages then things get progressively worse: it's not uncommon to see fiber optical pathways several times the length of the true distance between two cities or countries (!). If we check the global topology of undersea fiber optical cables:

Some places are very far off the beaten path of the Internet, and for example there's no direct routes from Asia to Europe or from Australia to Europe. There are very few straight lines - which adds extra hops and distance to over the ~30% slowdown that light moving in Corning glass already causes.

Now as destinations get closer to each other the latency of fiber optical networks improves: so for example two cities that are 100 km apart usually are going to have better latencies than LEO constellation - except if the Internet backbone center of the country is hundred of kms away, which is frequently the case.

Not to mention the incremental cost: a single intercontinental cable takes years to plan, lay and costs hundreds of millions of dollars overall. A LEO constellation connects everyone to everyone within the constellation's latitude limits.

But in terms of global backbone of Internet traffic, a LEO constellation is superior in almost every regard to fiber optical networks, and they are much easier and lower cost to scale up in capacity as well - so they will almost certainly win the race for more and more bandwidth.

 

[neroden]

Sorry, incorrect. The speed of light in fiber is about 1/3rd speed of light in a vacuum.

Tha atmosphere isn't a vacuum. And of course the distance is shorter on a *direct* fiber link than on a bounce through space. And there's more radio interference in the air than in a fiber cable. So there are a lot of factors you have to figure in when calculating latency. Wall Street high-speed traders will still co-locate next door to the stock exchanges with direct fiber links.

What matters for most purposes, however, is none of this -- what matters is the number of times the message has to be "reprocessed" at a router or switch (the number of hops) and the congestion on the route. For these things, Starlink should be very superior to almost everything due to the extremely large number of satellites which can communicate to each other. The minimization of the hop count is a big deal.

 

[neroden]

Big step will be once it can act on road signs. Stop signs, red lights, priorities etc.

Then it will also be how to navigate in parking lots where there are no clear navigation signs.

Parking lots with marked spaces are step one; parking lots WITHOUT markings are the second and harder part of the parking lot problem; parking lots made of grass (such as used for festivals) are the third and very hard part.

 

[Fact Checking]

Fine, but that's not level 5 "full self driving".

Where do you get the 'level 5' from? Tesla FSD aims for Level 4 autonomy with eventual Level 5 autonomy (much) later on.

 

[Unpilot]

It doesn't. In this case, I think Musk is right; the whole point of these satellites is that they're cheap (i.e. if they break, put up more
); the test-launched satellites work, and going through extra rounds of testing is simply wasting time. What this will affect is Comcast, Verizon, et al. I'll be buying that "pizza box" satellite connection as soon as I can. It'll have a waiting list longer than the solar roof, possibly longer than Model 3.

I believe this is correct. How many people would ditch Comcast or ATT in a heartbeat if there was a viable alternative.
I would...in other words GET IT UP THERE AND TAKE MY CASH!

 

[Fact Checking]

I'll be buying that "pizza box" satellite connection as soon as I can. It'll have a waiting list longer than the solar roof, possibly longer than Model 3.

Same here: having an early slot in the geographically limited number of Starlink subscriptions is going to be worth quite a bit a few years after introduction. Starlink satellites will be able to form narrow down-link beams only so much, which means they'll have to manage channels by their frequency slots which are limited, i.e. there will be a hard limit on subscriber density. This will limit the number of maximum possible Starlink subscribers in metro areas and in their ~100 km surroundings.

I don't know how good their down-link beam-forming is going to be, but if it's say a 5 km diameter circle/ellipsoid (which would be very good beam-forming already), that still limits high population density areas to a couple of dozen subscribers per square kilometer - possibly less, at least with the initial constellation.

 

[neroden]

I read that to mean that even though Tesla did not provide the capital for the equipment, nor does it lease the equipment, GAAP rules in ASC 840 nevertheless require that because they agreed to purchase all production from said equipment they must, for accounting purposes, pretend that they have leased the equipment even though they have not. IANAA so I could be totally wrong.

It means that it's basically accounted for as if they owned it, and borrowed money to buy it. Although it's technically leased.

An awful lot of leases are, in economic substance, purchases with financing; they're now being called "financing leases". This is sensible accounting.