News for hackers. Stuff that matters.
Re: Google announces a new generation for its TPU machine le... by sabalaba on Tuesday May 08, 2018 @03:42PM (#17024162)
This is great news. It’s extremely important to the research community that large companies enter into the DL silicon space to contend with NVIDIA’s monopoly.

NVIDIA is now exerting pricing power to the point where they’ve decided to train their sales people to disregard the metric that is most important to customers: cost of training. Talk with one of their enterprise sales people and you’ll find they’ll say things like “FLOPS / $ doesn’t matter” to justify a 10x increase in price for their TESLA line. As history has shown, a monopolist sows the seeds of their own destruction and, by disregarding the metrics that matter, they alienate their customers.

Here are open source projects that you can contribute to to break the monopoly:

ROCm: https://rocm.github.io/

MIOpen: https://github.com/ROCmSoftwarePlatform/MIOpen

TensorFlow: https://github.com/tensorflow/tensorflow

Re: Google announces a new generation for its TPU machine le... by twtw on Tuesday May 08, 2018 @04:16PM (#17024557)
Instead of being able to buy your own nvidia gpu and run any cuda or opencl on it, now you can run only tensor flow on a tpu only in Google cloud.

How fantastic.

Re: Google announces a new generation for its TPU machine le... by sabalaba on Tuesday May 08, 2018 @05:01PM (#17025053)
Choice in the marketplace is still choice. And, Google might eventually decide to sell this to outside companies. Eventually the desire to build competitive advantages for Google Cloud might be surpassed by the desire to compete with NVIDIA. Commoditizing the complement is a powerful strategy.
Re: Google announces a new generation for its TPU machine le... by throwaway2048 on Tuesday May 08, 2018 @05:01PM (#17025056)
Yeah seems like a way more abusive position to me frankly, a Google monopoly would be way worse than a NVIDIA one.
Re: Google announces a new generation for its TPU machine le... by jacksmith21006 on Tuesday May 08, 2018 @06:50PM (#17026034)
Not a Google monopoly but instead another choice. To push Nvidia and others.
Re: Google announces a new generation for its TPU machine le... by chewxy on Tuesday May 08, 2018 @04:44PM (#17024864)
And only using Tensorflow.
Re: Google announces a new generation for its TPU machine le... by jacksmith21006 on Tuesday May 08, 2018 @05:10PM (#17025132)
It is about choice and driving down cost. Love google keeps pushing the envelope which is good for everyone
Re: Google announces a new generation for its TPU machine le... by hedgehog on Tuesday May 08, 2018 @04:14PM (#17024516)
We're working to provide more choice with PlaidML as well. You can use it to run Keras, ONNX, and TF models on AMD, Intel, and mobile/embedded GPUs, on most operating systems and OpenCL, LLVM, OpenGL, and soon CUDA and Metal, etc. We would welcome contributions for TensorFlow integration, ROCm support, and anything else that widens support to more people.

https://github.com/plaidml/plaidml

Re: Google announces a new generation for its TPU machine le... by floatboth on Tuesday May 08, 2018 @05:04PM (#17025083)
Cool! Are you considering Vulkan?
Re: Google announces a new generation for its TPU machine le... by hedgehog on Tuesday May 08, 2018 @09:48PM (#17026876)
A couple people internally have looked at Vulkan, it seems like it should work really well but we're not actively working on it. You can get an idea of how Vulkan support would look by looking at the existing HALs:

https://github.com/plaidml/plaidml/tree/master/tile/hal

It's kind of cool to look anyway because it's a lot of value in less code than you might expect (complexity is in the layers above HAL).

If someone wanted to add support for Vulkan or something more ambitious like Qualcomm Hexagon or the RPi with QPU that would be pretty cool though.

Re: Google announces a new generation for its TPU machine le... by tehsauce on Tuesday May 08, 2018 @04:17PM (#17024564)
TPUs seem like great hardware. Unfortunately google's insistence on keeping them in a walled garden is a major deterrent to building something that relies on their platform.
Re: Google announces a new generation for its TPU machine le... by jacksmith21006 on Tuesday May 08, 2018 @06:50PM (#17026040)
What makes sense. Would guess each cloud provider will do their own.
Re: Google announces a new generation for its TPU machine le... by shaklee3 on Wednesday May 09, 2018 @06:05PM (#17034588)
You seem to be missing the point again and again every time this topic comes up. There is a reason why TensorFlow is generic enough to run on many types of hardware. That allows people to port their code around. But there exists a world outside of tensorflow that you might not be aware of, and for that world this chip is not a competitor. That world is significantly larger than the TF world.
Re: Google announces a new generation for its TPU machine le... by kuschku on Tuesday May 08, 2018 @07:14PM (#17026189)
And that helps me get one in my own servers how?
Re: Google announces a new generation for its TPU machine le... by jacksmith21006 on Wednesday May 09, 2018 @08:30AM (#17029285)
Common library above is how. Top one today is TF.
Re: Google announces a new generation for its TPU machine le... by kuschku on Wednesday May 09, 2018 @03:22PM (#17033010)
So how does that get me a TPU in my own servers?
Re: Google announces a new generation for its TPU machine le... by jacksmith21006 on Thursday May 10, 2018 @09:24AM (#17038594)
Does not make much sense to buy your own chips any longer.

TF removes the silicon from the equation. So if Nvidia cheaper use but if TPUs cheaper use. If something new that is cheaper then move. TF is the key.

Re: Google announces a new generation for its TPU machine le... by zitterbewegung on Tuesday May 08, 2018 @03:48PM (#17024207)
Well until that actually happens I will be buying more NVIDIA stock.

I have done deep learning also and I don't see NVIDIA being an incumbent that is easy to be knocked out. I think we would need some distruption such as a new technique that is more performant or different design goals. Intel is still leading the way and I don't see NVIDIA being unseated with its head start anytime soon.

Re: Google announces a new generation for its TPU machine le... by sabalaba on Tuesday May 08, 2018 @03:53PM (#17024265)
I'm not as pessimistic. 3dfx and Matrox didn't look like easy to knock out opponents in the 90s and yet here we are. Markets are powerful incentive mechanisms.

"Slowly grinds the mill of the gods, but it grinds fine."

Re: Google announces a new generation for its TPU machine le... by throwaway2048 on Tuesday May 08, 2018 @05:03PM (#17025072)
Costs to compete in the silicon business have inflated massively since that era, its perhaps 100 or even 1000 times more expensive to build stuff like fabs (although NVIDIA dosent own fabs), and design chips.
Re: Google announces a new generation for its TPU machine le... by jacksmith21006 on Tuesday May 08, 2018 @06:56PM (#17026072)
But this makes it so Google is not dependent on Nvidia. But also push Nvdia.
Re: Google announces a new generation for its TPU machine le... by ksec on Wednesday May 09, 2018 @04:51AM (#17028324)
>3dfx and Matrox didn't look like easy to knock out opponents in the 90s and yet here we are.

Matrox didn't give much of a damn to 3D gaming, 3Dfx were bust doing all the wrong things. It was certainly not "easy", but don't seems hard or impossible either at the time.

Re: Google announces a new generation for its TPU machine le... by zitterbewegung on Tuesday May 08, 2018 @10:10PM (#17026965)
I will also change my Cable provider to something other than Comcast because I hate their service. Oh wait they are the only provider in my area...
Re: Google announces a new generation for its TPU machine le... by beagle3 on Tuesday May 08, 2018 @05:25PM (#17025302)
An up-and-coming project in an up-and-coming language:

ArrayMancer: https://mratsim.github.io/Arraymancer/

Written in Nim, which looks and feels a lot like Python, but runs a lot like C++ or Rust, especially in the speed and FFI departments, and deploys like Go as a small single fast executable. Already supports CPU, CUDA and OpenCL. Still early, but with a lot of promise. Use is modeled on some mixture of PyTorch and Keras.

Re: Google announces a new generation for its TPU machine le... by chewxy on Tuesday May 08, 2018 @05:50PM (#17025528)
Self-plug: Gorgonia[0] can support OpenCL as well (but time and effort is required to write support libs). Would appreciate contributions. I recently gave a talk on how to interface with foreign libs[1]. Writeup soon to come.

[0]: https://github.com/gorgonia

[1]: https://www.youtube.com/watch?v=7R39CMW4MJQ

Re: Google announces a new generation for its TPU machine le... by bitL on Tuesday May 08, 2018 @07:25PM (#17026243)
I think NVidia was milking the cow in anticipation of moves from Google, AMD and MS in their space, leading to a pricing war.
Re: Google announces a new generation for its TPU machine le... by punnerud on Tuesday May 08, 2018 @05:56PM (#17025593)
Water cooling on their TPU is probably because they cost way more to buy than the extra cost of special cooling. It is also probably to increase “spec” on their new version without the best improvements in hardware/silicon.