AlphaZero

[Damir]

Can be found here: https://github.com/Zeta36/chess-alpha-zero

Good Luck All !!

[Milos]

Can be found here: https://github.com/Zeta36/chess-alpha-zero

Good Luck All !!

That has nothing to do with DeepMind publicity stunt.
Srdja should actually better think about copyright issues since he is clearly misinforming ppl.

[Damir]

It was supposed to be a joke, nothing serious. No reason to get upset Milos.

[Damir]

If you click on Zeta 36 profile it says his name is Samuel Graván. He is not related to Srdja in any way.

[Milos]

It was supposed to be a joke, nothing serious. No reason to get upset Milos.

No I don't mind you Damir, but as a friendly advice to Srdja, one thing is to try to do things as described in the paper, but totally other to take the same name that Google used. That is almost certainly a copyright infringement even though AlphaZero is not yet a product.

[Milos]

If you click on Zeta 36 profile it says his name is Samuel Graván. He is not related to Srdja in any way.

Ok, I thought it is Srdja Matovic coz he had zeta profile on github, but it's another guy. Anyway, possible copyright infringement still stands.

[Damir]

Yes copyright infringment still stands, but not unless the guy is on the google team, and wanted to share how Alphazero mind really works.. But if it is other party using google's name, than that could be a problem, as you say...

[CheckersGuy]

It was supposed to be a joke, nothing serious. No reason to get upset Milos.

No I don't mind you Damir, but as a friendly advice to Srdja, one thing is to try to do things as described in the paper, but totally other to take the same name that Google used. That is almost certainly a copyright infringement even though AlphaZero is not yet a product.

True. One shouldn`t misinform ppl like that. Btw, I don`t thiink Alphago nor alphaZero will be commercial products. Demis Hassibs stated in an interview that there will be plenty of ppl that will copy their approach once they publish their papers. The statement was about alphago but I suspect that it isnt any different with alphaZero

[hgm]

There would't be much marker for AlphaZero, as you need a TPU to usefully run it.

However, now that NN are proving their worth in various applications, it becomes a good question whether in the future we will see more NNs in common applications, and TPU technology seeping into our PCs and phones to efficiently run those. There has been a time where floatig-poit arithmetic was considered so difficult and outladish that standard CPUs did not support them, and you needed a separate FPU chip (the 8087) to equip your PC with that capability. Now we have them as a standard feature in every CPU core. It is quite possible that TPUs will go the same way.

Note that the Google TPUs are not really much more complex in terms of number of transistors or power consumption than state-of-the-art CPU chips with many x86 cores. (And they are pretty big, as TPUs go.) The benefit of going from 4 cores to 8 cores might be very small for the average user, while having 4 CPU cores plus a TPU on that chip could open a plethora of new applications.

[Milos]

There would't be much marker for AlphaZero, as you need a TPU to usefully run it.

However, now that NN are proving their worth in various applications, it becomes a good question whether in the future we will see more NNs in common applications, and TPU technology seeping into our PCs and phones to efficiently run those. There has been a time where floatig-poit arithmetic was considered so difficult and outladish that standard CPUs did not support them, and you needed a separate FPU chip (the 8087) to equip your PC with that capability. Now we have them as a standard feature in every CPU core. It is quite possible that TPUs will go the same way.

Note that the Google TPUs are not really much more complex in terms of number of transistors or power consumption than state-of-the-art CPU chips with many x86 cores. (And they are pretty big, as TPUs go.) The benefit of going from 4 cores to 8 cores might be very small for the average user, while having 4 CPU cores plus a TPU on that chip could open a plethora of new applications.

That is all true, but Google doesn't have much saying into this. It is all on NVIDIA. In terms of chip design TPU is totally ordinary thing, 22nm process is already 2-3 years behind cutting-edge, they most probably used old TSMC process.
Once NVIDIA actually concludes that dedicated ML chips are good for business we'll have ML chip of 10x TPU performance in no time.
However, atm I'm pretty sceptical that even 15% of for GV100 would be used for ML. It would still be 60% gaming and 25% mining. So no market yet for dedicated chip.