benchmark new smartphones:

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Rebel, chrisw

Chessqueen
Posts: 5663
Joined: Wed Sep 05, 2018 2:16 am
Location: Moving
Full name: Jorge Picado

Re: benchmark new smartphones:

Post by Chessqueen »

AdminX wrote: Thu Jul 14, 2022 12:07 am Just ordered Pixel 6 Pro, plan to 'De-Google' it with 'GrapheneOS' and hope that will not cause to many problems with my chess apps. Has anyone benchmarked this with say Stockfish?
I have a Pixel 6 Pro, also an Aple Iphone 13 Pro Max, Where is the Benchmarch ?
I am ready to swap brain with Stockfish chip. https://www.facebook.com/share/iHzQpSjc ... tid=oFDknk
Chessqueen
Posts: 5663
Joined: Wed Sep 05, 2018 2:16 am
Location: Moving
Full name: Jorge Picado

Re: benchmark new smartphones:

Post by Chessqueen »

Chessqueen wrote: Thu Jul 14, 2022 8:25 am
AdminX wrote: Thu Jul 14, 2022 12:07 am Just ordered Pixel 6 Pro, plan to 'De-Google' it with 'GrapheneOS' and hope that will not cause to many problems with my chess apps. Has anyone benchmarked this with say Stockfish?
I have a Pixel 6 Pro, also an Aple Iphone 13 Pro Max, Where is the Benchmarch ?
If it is this position after 2 minutes of analysis 74408 nps and 739 k with Pixel 6 Pro fxe6 best move, DroidFish Stockfish 15 NNUE
[d]1q6/1r2k1p1/4pp1p/1P1b1P2/3Q4/7P/4B1P1/2R3K1 w
I am ready to swap brain with Stockfish chip. https://www.facebook.com/share/iHzQpSjc ... tid=oFDknk
h1a8
Posts: 512
Joined: Fri Jun 04, 2010 7:23 am

Re: benchmark new smartphones:

Post by h1a8 »

Are there any ios apps that run stockfish NNUE (14 or newer) using ALL CORES and at FULL SPEED?

Smallfish hides the kn/s and doesn't seem to reach high depths fast enough.

Chess Studio shows a very low 340kn/s (both with and without NNUE enabled) on iphone 14 pro max

Other apps I've tried get from 2000kn/s to 4000kn/s on iPhone 14pro max, depending on NNUE on or off.

The original stockfish app seems to utilize all the cores at full speed but it's using a very old stockfish engine.
I would gladly pay (donate) to anyone who can update the original stockfish app with newer engine.
User avatar
towforce
Posts: 11706
Joined: Thu Mar 09, 2006 12:57 am
Location: Birmingham UK

Re: benchmark new smartphones:

Post by towforce »

I stand by my suggestion made in this thread on 01-Aug-2020: a device that gets a good score on standard performance metrics is likely to run your chess program quickly. For example, if you want to know a particular phone's Geekbench score, Google it, and 2 seconds later you'll know it. If you just want the fastest phone, you'll find plenty of Geekbench league tables.

My guess is that the correlation between Geekbench scores and specialist chess software scores will be higher than 0.9: if so, the Geekbench scores will be the way to go: they're everywhere, and more or less every device will have been tested.

tbf, this won't answer h1a8's question above.
The simple reveals itself after the complex has been exhausted.
User avatar
mclane
Posts: 18769
Joined: Thu Mar 09, 2006 6:40 pm
Location: US of Europe, germany
Full name: Thorsten Czub

Re: benchmark new smartphones:

Post by mclane »

I have put google services on my google smartphone with harmony 4 OS.
My p60 pro runs very good now.
And uses less ernergy then before.
What seems like a fairy tale today may be reality tomorrow.
Here we have a fairy tale of the day after tomorrow....
h1a8
Posts: 512
Joined: Fri Jun 04, 2010 7:23 am

Re: benchmark new smartphones:

Post by h1a8 »

towforce wrote: Sat May 11, 2024 5:54 pm I stand by my suggestion made in this thread on 01-Aug-2020: a device that gets a good score on standard performance metrics is likely to run your chess program quickly. For example, if you want to know a particular phone's Geekbench score, Google it, and 2 seconds later you'll know it. If you just want the fastest phone, you'll find plenty of Geekbench league tables.

My guess is that the correlation between Geekbench scores and specialist chess software scores will be higher than 0.9: if so, the Geekbench scores will be the way to go: they're everywhere, and more or less every device will have been tested.

tbf, this won't answer h1a8's question above.
Geekbench integer scores highly correlate with kn/s
PROVIDED THE CHESS PROGRAM IS USING ALL CORES AT FULL SPEED. Some programs use less than half the available cores for stockfish (chess studio for example)