We already have 2 of those in Belgium. No need to wait.By the time Tesla gets around to making their European GF they will surely have a choice of several former car production sites, that can be had for a good price.
You can install our site as a web app on your iOS device by utilizing the Add to Home Screen feature in Safari. Please see this thread for more details on this.
Note: This feature may not be available in some browsers.
We already have 2 of those in Belgium. No need to wait.By the time Tesla gets around to making their European GF they will surely have a choice of several former car production sites, that can be had for a good price.
Netlix is an enigma to me. Too hard to predict where it's headed in 5, 10 years. I'm not comfortable investing in it at this valuation. But I agree with your perspective. Essentially it's about scalability. An inherent advantage of tech, and esp Netflix.In my view that kind of fund managers are unlikely to change their view. Based on his statement, he doesn't understand those companies at all. If he can't understand Tesla, Netflix, BMW in the first place, he won't understand them later. I think his fund will be wiped out at some point.
To give you an extreme example, if a company holds 100 ton of gold, meanwhile the company does some gold mining, made a little bit of money by the year end because the mining activity was very costly, P/E is 200. This manager would view that as very expensive. They were taught a P/E of 10~15 is more appropriate. Netflix has large amount of content that people will continue to pay $10 a month to subscribe, and they are ready to expend into china. Their subscription could double to $20B a year. They also could monetize on these 200 million subscribers in the future, in addition to the monthly fee... Not saying NFLX is a good buy now, I'm saying that manager doesn't know what he is doing.
Edit: the manager talked about Netflix P/E is 200. Does he understand that's because Netflix is using all the revenue to produce more content, which is the right thing to do. He should have shorted Amazon in the past, because Amazon never had good earnings.
You can, I'm just giving my opinion that it doesn't really matter.
All I hold are Feb 19 and Aug 19 calls. No shares. I just see other companies' Q3 earnings reports as distractions at best. WE are the ones cooking with oil right now.It matters to me.
I have shares, Aug19 calls, a boatload of late Nov18 calls, and some weeklies.
I’m equally concerned about short term movements and long term movements.
I believe the clock starts ticking now on the land auction in Shanghai. It’s early morning in china and I believe if Tesla is the sole bidder for the plot then the announcement can come as soon as ..today China time.
Wonder if this impacts SP.
Securing land in a zone that both has no tariffs and qualifies for Chinese government rebates while remaining solely owned by Tesla is a tremendous competitive advantage in the largest market on the planet.
Even if Tesla get the land, that's not a "tremendous competitive advantage", I can only say potentially removed a "tremendous competitive disadvantage".
while remaining solely owned by Tesla
Is the computer being replaced with ARM, or X64? If ARM, is it Tesla grown or an existing solution?
I don't know NN use case specifically but HBM might be overkill for Tesla's purpose (though I am aware that Google is going with HBM on some of their TPU designs). If it does come with HBM (and thus likely interposer packaging) that would explain Elon's comment about it costing the same as the old Nvidia hardware. HBM is expensive right now, and interposer's aren't free.HBM and interposer.
There's no need for SOI. The TDP shouldn't be a problem, and plenty of automotive grade IC get by fine on just about any process that isn't garbage. Especially on Model 3, since it's liquid cooled (S/X might be air cooled? I'm less familiar with them), TDP and operating temp range should be a non issue. As for being large - this seems unlikely. Being specialized, I doubt that even at it's much higher performance than the old Pascal GPU architecture for NN processing, that it is larger than Pascal. It likely yields better due to smaller size (meaning for a given average defect rate per wafer, more 100% functional parts).Or the Tesla AI chip has a monster size and/or low yields. Given that this thing has to run in a harsh automotive environment and requires a combination of high performance and low power, it may be an SOI chip, which is also more expensive.
The Parker chips contain a Pascal GPU: Introducing Parker, NVIDIA’s Newest SOC for Autonomous Vehicles | NVIDIA Blog
While it appears that the GP102 has 3328 CUDA cores.
BTW., if they are really using a GP102-equivalent Pascal chip with 3,328 CUDA cores, then that's about 85% of the available GPU computing power (the two Parker chips have 256/256 CUDA cores) - so I think the discrete Pascal GPU is probably processing most of the cameras. The 7% CUDA cores available on the Parker side might be able to process down-scaled images only, with reduced size neural nets.
I.e. exactly as you originally suggested.
Yeah, so visual input processing is the most computing intense part of full self-driving. Tesla has 8 cameras, and if you want to process each at 100 fps (one frame every 10 milliseconds), at the native HD resolution of the cameras, that's a lot of processing.
Right now they process everything, all frames from all 8 cameras with a single discrete GPU I believe, on an Nvidia GP102 based board.
But now that they have their own discrete NN chip, the Tesla AI chip, in future iterations (HW4, HW5) they could use the following computer topology within the board, with very little additional cost (the AI chips probably cost only a few dollars to make each - most of the cost is in making the board):
Code:[AI Chip #1] [AI Chip #2] \ / [GPU RAM] / \ [AI Chip #3] [AI Chip #4]
I.e. four chips and shared RAM of say 16 GB high-speed GPU RAM with multiple access channels so that all CPUs can use the RAM all the time without slowing down each other.
(There's also the question of whether the Tesla AI chip uses separate RAM modules - a possible alternate design would be for the RAM to be integrated into the AI chip itself, as a sort of very fast transistor based SRAM. This would have a number of other advantages as well, such as close proximity of NN 'weight' data with the functional units representing 'neuron' nodes.)
But assuming that RAM is separate from the chip, the above board layout is a possible topology, where Chip 1 would handle cameras 1-2, Chip 2 would handle cameras 3-4, etc. While not all cameras have the same pixel count, the processing overhead is still similar and scales with the complexity of their neural networks.
Note that this way the total computing throughput of the system can be increased by a factor of 2x, 4x and 8x with very little additional cost other than a higher power envelope.
I'm reasonably sure HW3 is going to feature one AI chip (they want to keep it simple initially, and it appears the chip is plenty fast already) - if it features two chips it will be for redundancy and fail-over perhaps, not to increase performance.
All of this is speculation though - I'm sure we'll hear more about the details once the HW3 release gets closer ...
"Chiplet" design is the future for affordable computing power. Just look at the pricing of the big CPUs from Intel and AMD now. AMD has up to 4 smaller dies, versus a single monolithic for Intel, and this means massively cheaper production costs as your yields are much higher with smaller dies. There are of course tradeoffs, but few if any of those matter for highly parallel tasks like processing NNs.They could also alternate between two chips on each frame, at a cost of 1 frame of delay if they get over the capacity of a single chip solution. I would prefer a new chip with twice the processing power, it could leave the rest of the system without changes.
I think an important but underappreciated advantage of the Tesla chip is the power advantage. GPU’s may use hundreds of Watts, while the Tesla chip may use a lot less. In an robotaxi situation where the chip runs 24/7 full tilt, the cost advantage in power may be much bigger than the cost of the GPU itself.
Here’s the latest part in my blog series on how Tesla can win the media war.
DaveT on Chatstarter: "Part 5: The Turning Tide - The Campaign Tesla Ought to Launch to Take Back Their Narrative"
I would still say media coverage is negative toward Tesla, but yes I do agree slightly more positive the last couple weeks with Model 3 sales growing rapidly and taking away sales from other auto manufacturers. But Tesla is far from out of the woods in regards to winning the media war. I think it's just a small respite right now.I assume you are aware of the already changing, more positive narrative, we are seeing lately in the media. I think the increased and pointed attacks on the negative bias in the media are actually paying off. I think the scope of what you are proposing is too large and to some degree unnecessary. Daily events for 30 journalists would be overkill, maybe once a week should be sufficient. Not to mention diverting Model 3's away from customers to allow journalists in every country a 2 day driving experience seems like a poor idea at this point in time.
Completely agree but I think baring any negative surprises the tide will continue to turn. A lot of people seem to be tired of the unfair negativity and are fighting back. I'm not against your idea of Tesla taking greater control of the narrative, just the scope.But Tesla is far from out of the woods in regards to winning the media war.
In case you hadn't noticed Donn has been banned, along with a few others I guess. It's been mercifully rational here recently, (outside of some fantasizing about Elon being an active TMC contributor).Really Donn?
So popular mechanics has a new article titled "In defence of Elon Musk". I spent around half an hour reading it and my perspective has changed on the significance of some of the more erratic parts of Elons last few months. Even though I always supported him and understood he was under immense pressure from many sides, I was somewhat concerned about what was going on with the pedo tweets and the lack of a formalised approach to "funding secured ". Honestly, the last few months now feels like a footnote in an incredible story that is about to unfold. Nothing more. Please read if you have the time.
In Defense of Elon Musk
Some of us here buy short term calls.With all due respect, who tf cares?
Tesla often disregards larger market moves, and are any of our timelines even short enough for this to matter?
When Tesla's OWN earnings report hits, we'll have a juicy ol' time, I reckon