benchmark new smartphones: (Old Thread)

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

Moderators: bob, hgm, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
Locked
Alexandr Novikov
Posts: 22
Joined: Tue Sep 15, 2015 6:21 pm

Re: iPad Pro with A9X-chip

Post by Alexandr Novikov » Wed Mar 23, 2016 11:23 am

I have Ipad pro, how I can check the speed in chess?

User avatar
yurikvelo
Posts: 547
Joined: Sat Dec 06, 2014 12:53 pm

Re: iPad Pro with A9X-chip

Post by yurikvelo » Wed Mar 23, 2016 12:20 pm

Alexandr Novikov wrote:I have Ipad pro, how I can check the speed in chess?
https://itunes.apple.com/gb/app/stockfi ... 58605?mt=8

Gurcan Uckardes
Posts: 196
Joined: Tue Oct 28, 2014 11:42 pm
Contact:

Re: iPad Pro with A9X-chip

Post by Gurcan Uckardes » Wed Mar 23, 2016 1:46 pm

Regarding Geekbench, i confirm the good correlation with chess performance. A closer look can also reveal that the "cpu integer" score is even more representative.
In chess, graphics don't matter, floating point either. Integers math and L1 cache decide!
My blog for Android users: http://chesstroid.blogspot.com

User avatar
towforce
Posts: 9862
Joined: Wed Mar 08, 2006 11:57 pm
Location: Birmingham UK

Re: iPad Pro with A9X-chip

Post by towforce » Wed Mar 23, 2016 3:36 pm

Gurcan Uckardes wrote:Regarding Geekbench, i confirm the good correlation with chess performance. A closer look can also reveal that the "cpu integer" score is even more representative.
In chess, graphics don't matter, floating point either. Integers math and L1 cache decide!
Thanks for this! 8-)
If you can't say something nice then don't say anything - link

User avatar
JuLieN
Posts: 2948
Joined: Mon May 05, 2008 10:16 am
Location: Nantes (France)
Contact:

Re: iPad Pro with A9X-chip

Post by JuLieN » Wed Mar 23, 2016 5:20 pm

@Paul :

Yes, the numbers are NPS. This is not a very scientific measurement, considering that testing Stockfish on a Galaxy S7 during one minute is probably like testing it for hours on a Palm Pre : to get a comparable result to the one on the Palm Pre, we should measure the NPS after a few seconds. Add to this that all those smartphones were tested with different versions of Stockfish...

Still, the results give a good overview of the mobile hardware's progress. The ARM CPU is not far from catching up with desktop CPUs, at a fraction of the power usage...

Also, I'm disappointed with the Snapdragon's results. I was expecting even higher results than with the Exynos. This is surprising. As has been mentioned, this probably has something to do with the size of the cache. Also, the Exynos has 8 cores.
"The only good bug is a dead bug." (Don Dailey)
Image [Blog: http://tinyurl.com/predateur ] [Facebook: http://tinyurl.com/fbpredateur ] [MacEngines: http://tinyurl.com/macengines ]

User avatar
mclane
Posts: 17208
Joined: Thu Mar 09, 2006 5:40 pm
Location: US of Europe, germany
Full name: Thorsten Czub
Contact:

Re: iPad Pro with A9X-chip

Post by mclane » Wed Mar 23, 2016 6:33 pm

Snapdragon Sucks for a long time. I remember when I had the note2.
It had 714.000 when I remember it right and meanwhile snapdragon were lousy. This has not changed over the years that have gone, the company can only make a deal because of the GPU. But the CPU is IMO lousy.
What seems like a fairy tale today may be reality tomorrow.
Here we have a fairy tale of the day after tomorrow....

User avatar
towforce
Posts: 9862
Joined: Wed Mar 08, 2006 11:57 pm
Location: Birmingham UK

Re: iPad Pro with A9X-chip

Post by towforce » Wed Mar 23, 2016 9:45 pm

JuLieN wrote:@Paul :

Yes, the numbers are NPS. This is not a very scientific measurement, considering that testing Stockfish on a Galaxy S7 during one minute is probably like testing it for hours on a Palm Pre : to get a comparable result to the one on the Palm Pre, we should measure the NPS after a few seconds. Add to this that all those smartphones were tested with different versions of Stockfish...

Still, the results give a good overview of the mobile hardware's progress. The ARM CPU is not far from catching up with desktop CPUs, at a fraction of the power usage...

Also, I'm disappointed with the Snapdragon's results. I was expecting even higher results than with the Exynos. This is surprising. As has been mentioned, this probably has something to do with the size of the cache. Also, the Exynos has 8 cores.
Thanks for this answer to my question. 8-)
If you can't say something nice then don't say anything - link

User avatar
mclane
Posts: 17208
Joined: Thu Mar 09, 2006 5:40 pm
Location: US of Europe, germany
Full name: Thorsten Czub
Contact:

Re: iPad Pro with A9X-chip

Post by mclane » Sun Apr 03, 2016 7:21 pm

Motorola X Style (6 cores SD808 64bit Android6) 1.524.000


moto X Style with Android 6.0, little better then
LG g3 with the same CPU.

Not sure which Android Version the LG G3 was benchmarked.


Interesting that the 808 6 core CPU has almost no throttling.
it begins with the same NPS and up to 2 minutes holds the value.
As if heat is something not known for the device.

IMO x style is known as x pure in USA.
What seems like a fairy tale today may be reality tomorrow.
Here we have a fairy tale of the day after tomorrow....

User avatar
mclane
Posts: 17208
Joined: Thu Mar 09, 2006 5:40 pm
Location: US of Europe, germany
Full name: Thorsten Czub
Contact:

Re: iPad Pro with A9X-chip

Post by mclane » Mon Apr 11, 2016 9:42 pm

Htc m10 seems to be a nice mobile phone.
Maybe it beats lg G5 , iPhone and Samsung s7 and Huawei p9.

Htc urgently needs a hit.
What seems like a fairy tale today may be reality tomorrow.
Here we have a fairy tale of the day after tomorrow....

Nay Lin Tun
Posts: 595
Joined: Mon Jan 16, 2012 5:34 am

Re: iPad Pro with A9X-chip

Post by Nay Lin Tun » Wed Apr 13, 2016 2:16 am

JuLieN wrote:@Paul :

Yes, the numbers are NPS. This is not a very scientific measurement, considering that testing Stockfish on a Galaxy S7 during one minute is probably like testing it for hours on a Palm Pre : to get a comparable result to the one on the Palm Pre, we should measure the NPS after a few seconds. Add to this that all those smartphones were tested with different versions of Stockfish...

Still, the results give a good overview of the mobile hardware's progress. The ARM CPU is not far from catching up with desktop CPUs, at a fraction of the power usage...

Also, I'm disappointed with the Snapdragon's results. I was expecting even higher results than with the Exynos. This is surprising. As has been mentioned, this probably has something to do with the size of the cache. Also, the Exynos has 8 cores.
What setting do you do for S7 edge? I got 2.2k to 2.5KNPS in starting position on droidfish. There is no difference between 4 cpus vs 8 cpus in my phone though the cpu is Exynos 8 cores. What is the the best setting for droidfish on S7edge?

Locked