4 ms·

Am I missing something, or are they not only not documenting the chip, they are also not even releasing the compiler, but requiring you to use a cloud based com

by lambda 8y ago

Am I missing something, or are they not only not documenting the chip, they are also not even releasing the compiler, but requiring you to use a cloud based compiler: > You need to create a quantized TensorFlow Lite model and then compile the model for compatibility with the Edge TPU. We will provide a cloud-based compiler tool that accepts your .tflite file and returns a version that's compatible with the Edge TPU. This seems like a new low in software freedom, and pretty risky to depend on as Google is known to shutter services pretty often and could just decide to turn off their cloud-based compiler at any time they feel like.

diminoten 8y ago
Google's reputation for shuttering services is vastly overstated, and beating this particular dead horse every time Google creates something new just wastes people's time.
- mlindner 8y ago
  Well they decided to not only make it closed source but also lock it up behind a http frontend so you can't even reverse engineer it. Criticizing Google for shuttering things suddenly that companies are depending on is rightly justified.
  - diminoten 8y ago
    Literally not what they do, but sure, keep up the absurd lie.
royal_ts 8y ago
Which paid Services are shutdown regularly?
- seanp2k2 8y ago
  There are even dedicated sites for the products they've killed off over the years: https://www.wordstream.com/articles/retired-google-projects https://www.wordstream.com/articles/retired-google-projects https://gcemetery.co/ https://gcemetery.co/ https://killedbygoogle.com/ https://killedbygoogle.com/ https://en.wikipedia.org/wiki/Category:Discontinued_Google_s https://en.wikipedia.org/wiki/Category:Discontinued_Google_s ervices https://www.businessinsider.com/discontinued-google-products-2016-8 https://www.businessinsider.com/discontinued-google-products...
  - ijpoijpoihpiuoh 8y ago
    Very few of which, if any, appear to be paid, which is what the parent poster is asking about. (Not sure if they edited that bit in after you responded?)
- 076ae80a-3c97-4 8y ago
  https://en.wikipedia.org/wiki/List_of_Google_products#Discontinued_products_and_services https://en.wikipedia.org/wiki/List_of_Google_products#Discon...
jacquesm 8y ago
Chips without a public toolchain are not worth investing your time in. It is bad enough if your work is tied to specific hardware for which there may at some point not be a replacement but to not even have the toolchain under your control makes it a negative.
- digitalzombie 8y ago
  Seriously Intel is already struggling after buying Nervana. I went to their shing dig and they were working their butt off to wow the developers who were invited. When I asked for hard number they were very mum about that and very evasive. The timeline for Nervana chip have been always seemingly in this mystical horizon that is never solidified to a real date but over yonder. Google is going to pull this crap? They got better software expertise than Intel though they may be able to do it. But after that fiasco with Angular 1 to 2 I wouldn't trust Google with any early version number.
  - Traster 8y ago
    Nervana had a lot of other issues. It was trying to produce an ASIC with 50 employees. When they got aquired by Intel the first step they had to tackle was hiring the engineers necessary to actually produce an ASIC which innevitably slows down production, and then on top of that they got caught in the Intel 10nm bear trap.
- narrator 8y ago
  AI is too powerful a technology to let it out there to the masses. People might use it for killer drones after all. All users of AI must be tightly controlled and registered with the authorities! This is the problem with certain kinds of technology that are bumping up against the edge of innovation. They're too powerful and if these technologies get in the hands of the DIY set, governments will lose control so they have to DRM and regulate everything. Heck, it's a problem with old technology. Many weapons aren't that complicated technologically, but their production and use are tightly regulated. Edit: I'm not saying this is a good thing, I'm just deconstructing their though process for tight control over AI tech going forward.
yalogin 8y ago
Google has mastered the art of using open source to crush competition like they did with Chrome and Android. They ever revealed their main moneymakers ever. For example they opened up their mapreduce technique after they have moved on from it.
- delroth 8y ago
  What is your source for Google having moved on from MapReduce in 2004 (date when the MR paper was published)?
dekhn 8y ago
it doesn't require a cloud based compiler; the quote above shows that you use TF-Lite, which is an open source project, or a cloud-based tool for people who don't want/need/have the ability to work with TF-Lite. [UPDATE] I misread and assumed the previous case (where no cloud tool was required) was still true (I worked with previous versions of this device).
- lambda 8y ago
  Do you have a source for this, or are you just reading the statement I quoted differently than I am? The way I read the quote, you use TF-Lite to produce a quantized TF-Lite model, and then use a cloud based compiler to compile it for the actual chip. This is why I asked "am I missing something." Do you have a reference for where the compiler exists in the open source TensorFlow project? Mostly, what I'm interested in is learning what capabilities their TPU provides, to see if it would be useful for other similar kinds of kernels like DSP (which, like machine learning kernels, also involves a lot of convolution). So I'm interested in looking at what the capabilities of the chip are, seeing what could be compiled to it. But I haven't found those docs, or found a compiler that could be studied. But maybe I'm not looking in the right place. Here's an overview of the architecture of their Cloud TPUs, which has some good architectural details but doesn't documet the instruction set: https://cloud.google.com/blog/products/gcp/an-in-depth-look-at-googles-first-tensor-processing-unit-tpu https://cloud.google.com/blog/products/gcp/an-in-depth-look-...
- xkapastel 8y ago
  It does require a cloud based compiler, the accelerator doesn't run TF-Lite models directly. The cloud based tool is what gives you the actual executable.

by lambda 8y ago

diminoten 8y ago

Google's reputation for shuttering services is vastly overstated, and beating this particular dead horse every time Google creates something new just wastes people's time.

mlindner 8y ago
Well they decided to not only make it closed source but also lock it up behind a http frontend so you can't even reverse engineer it. Criticizing Google for shuttering things suddenly that companies are depending on is rightly justified.
- diminoten 8y ago
  Literally not what they do, but sure, keep up the absurd lie.

royal_ts 8y ago

Which paid Services are shutdown regularly?

seanp2k2 8y ago
There are even dedicated sites for the products they've killed off over the years: https://www.wordstream.com/articles/retired-google-projects https://www.wordstream.com/articles/retired-google-projects https://gcemetery.co/ https://gcemetery.co/ https://killedbygoogle.com/ https://killedbygoogle.com/ https://en.wikipedia.org/wiki/Category:Discontinued_Google_s https://en.wikipedia.org/wiki/Category:Discontinued_Google_s ervices https://www.businessinsider.com/discontinued-google-products-2016-8 https://www.businessinsider.com/discontinued-google-products...
- ijpoijpoihpiuoh 8y ago
  Very few of which, if any, appear to be paid, which is what the parent poster is asking about. (Not sure if they edited that bit in after you responded?)
076ae80a-3c97-4 8y ago
https://en.wikipedia.org/wiki/List_of_Google_products#Discontinued_products_and_services https://en.wikipedia.org/wiki/List_of_Google_products#Discon...

jacquesm 8y ago

Chips without a public toolchain are not worth investing your time in. It is bad enough if your work is tied to specific hardware for which there may at some point not be a replacement but to not even have the toolchain under your control makes it a negative.

digitalzombie 8y ago
Seriously Intel is already struggling after buying Nervana. I went to their shing dig and they were working their butt off to wow the developers who were invited. When I asked for hard number they were very mum about that and very evasive. The timeline for Nervana chip have been always seemingly in this mystical horizon that is never solidified to a real date but over yonder. Google is going to pull this crap? They got better software expertise than Intel though they may be able to do it. But after that fiasco with Angular 1 to 2 I wouldn't trust Google with any early version number.
- Traster 8y ago
  Nervana had a lot of other issues. It was trying to produce an ASIC with 50 employees. When they got aquired by Intel the first step they had to tackle was hiring the engineers necessary to actually produce an ASIC which innevitably slows down production, and then on top of that they got caught in the Intel 10nm bear trap.
narrator 8y ago
AI is too powerful a technology to let it out there to the masses. People might use it for killer drones after all. All users of AI must be tightly controlled and registered with the authorities! This is the problem with certain kinds of technology that are bumping up against the edge of innovation. They're too powerful and if these technologies get in the hands of the DIY set, governments will lose control so they have to DRM and regulate everything. Heck, it's a problem with old technology. Many weapons aren't that complicated technologically, but their production and use are tightly regulated. Edit: I'm not saying this is a good thing, I'm just deconstructing their though process for tight control over AI tech going forward.

yalogin 8y ago

Google has mastered the art of using open source to crush competition like they did with Chrome and Android. They ever revealed their main moneymakers ever. For example they opened up their mapreduce technique after they have moved on from it.

delroth 8y ago
What is your source for Google having moved on from MapReduce in 2004 (date when the MR paper was published)?

dekhn 8y ago

it doesn't require a cloud based compiler; the quote above shows that you use TF-Lite, which is an open source project, or a cloud-based tool for people who don't want/need/have the ability to work with TF-Lite. [UPDATE] I misread and assumed the previous case (where no cloud tool was required) was still true (I worked with previous versions of this device).

lambda 8y ago
Do you have a source for this, or are you just reading the statement I quoted differently than I am? The way I read the quote, you use TF-Lite to produce a quantized TF-Lite model, and then use a cloud based compiler to compile it for the actual chip. This is why I asked "am I missing something." Do you have a reference for where the compiler exists in the open source TensorFlow project? Mostly, what I'm interested in is learning what capabilities their TPU provides, to see if it would be useful for other similar kinds of kernels like DSP (which, like machine learning kernels, also involves a lot of convolution). So I'm interested in looking at what the capabilities of the chip are, seeing what could be compiled to it. But I haven't found those docs, or found a compiler that could be studied. But maybe I'm not looking in the right place. Here's an overview of the architecture of their Cloud TPUs, which has some good architectural details but doesn't documet the instruction set: https://cloud.google.com/blog/products/gcp/an-in-depth-look-at-googles-first-tensor-processing-unit-tpu https://cloud.google.com/blog/products/gcp/an-in-depth-look-...
xkapastel 8y ago
It does require a cloud based compiler, the accelerator doesn't run TF-Lite models directly. The cloud based tool is what gives you the actual executable.