croman
Well-Known Member
Here's the TPU paper for anyone who wants to geek out over the possibilities: https://arxiv.org/pdf/1704.04760.pdf
Can we band together and get you something nice? Wine? A Tesla accessory?
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.
Here's the TPU paper for anyone who wants to geek out over the possibilities: https://arxiv.org/pdf/1704.04760.pdf
Are the Google TPUs generalized or optimized for a specific function? I.e. if they're either generalized or optimized for training (I forget the correct term) NNs, versus running already trained NNs, then the Tesla design may differ as it will surely be optimized only for running already compiled NNs.
Can we band together and get you something nice? Wine? A Tesla accessory?
Thanks for the informative replies/posts. I asked the question because I lack the necessary understanding of NN related tech to easily find the answer from the document, though I'm not surprised it was in there. It's an area of tech I haven't yet taken the time to become familiar with.They are optimized for performing matrix multiplication on 8 bit quantized neural networks who's mean dimension is over 256. This means they will be pretty good at running inference on the large majority of vision NN algorithms in use today.
The first version of the TPU (as detailed in the paper I linked above) is not appropriate for training networks with any extant algorithm - it's meant for deploying trained and pre-optimized networks. It happens to be well suited to running networks just like the ones that Tesla is currently using in the vehicle (TensorRT style optimized 8bit inference). Frankly, the TPU V1 itself would probably be a great chip for Tesla to deploy but for the fact that it's not commercially available. There are no commercial systolic array chips that I'm aware of. Google is releasing (this month) a chip of their design which they call the edge TPU which might be a systolic array and which could conceivably be useful in self driving applications - but I've not yet see the architecture or any detailed specs for it.
Edge TPU - Run Inference at the Edge | Edge TPU | Google Cloud
Edge TPU Devices
Of course there would be a bunch of small hardware and software changes, and Tesla would have to develop a set of software tools to enable efficient use of the new chip.
Second is the audaciousness of the architecture - it's ground breaking.
How much would you say the Autopilot neural network(s) have increased in size between the release right before v9 and the original networks going back to late 2016/early 2017?
The original network from early 2017 was a little toy demo they downloaded from the Internet.
That is almost not an exaggeration. But it is an exaggeration. But almost not.
It is not an exaggeration. It is just your opinion.
2) The front half of the NN in AP2 is basically Googlenet with a few notable differences:
- The input is 416x640 (original Googlenet was 224x224)
- The working frame size in Googlenet is reduced by 1/2 in each dimension between each of the 5 major blocks. The AP network omits the reduction between blocks 4 and 5 so that the final set of features is 2x2 times larger than in Googlenet.
[...]
After my first look at the version 40 NN I was surprised at how simple it was, conceptually, and how 'old' the circa 2015 architectural concepts were and speculated that perhaps this version of EAP was not getting much effort. (In the deep learning world 2 years is an eternity).
For perspective, V9 camera network is 10x larger and requires 200x more computation when compared to Google’s Inception V1 network from which V9 gets it’s underlying architectural concept.
This is not a matter of opinion, though I am certainly putting my own spin on it which is where the exaggeration comes from. But go back to the first page of this thread, where you will find @jimmy_d 's findings, which I willcherry-pickexcerpt here:
Like I said, I'm exaggerating and spinning, but the fact is that the NN in use in early 2017 was simplistic, already outdated, with an architecture cribbed from Googlenet (which was simplistic and outdated already at the time). This is something they threw together as quickly as possible to begin enabling the most simplistic AP features.
But it is your opinion, because you have no idea what Tesla used to create their FSD video. Nobody outside of a Tesla NDA knows what software was used to make that film.
NN Changes in V9 (2018.39.7)
Just wanted to echo our appreciation for your contribution, and ask:
Is developing a set of software tools a huge undertaking? What would the time frame be on this? And would this fall under Stuart Bowers responsibilities?
Here, do you mean the design of the architecture is new and groundbreaking, or that its design is close to an existing architecture but it's way bigger?
Oh, and you think the network has 5x more weights, but what about layers? Any increase in layers?
Amazing how you figure all of this stuff out without having access to the source code of this system!