This Antutu 6 benchmark seems very good, at least the total score. The SF nps seem very bad, though. Not sure what is this, could you re-run in Chess for Android? If this holds, and there is no another OS issue or bloatware, it would mean "stay away from Snapdragon 820".royb wrote:
OK, I think I found the problem with AnTuTu performance. I had turned on Developer extensions. When I switched them off and re-ran AnTuTu 6.0.1 I got these results:
Score: 134867
Device: SM-G935V
3D: 57356
UX: 40318
CPU: 31106
RAM: 6087
No huge difference though in Droidfish running the SF 7 engine with 4-threads: 1418K nodes per second. I am beginning to think that Droidfish has some issue that's not obvious other than not performing nearly as well as would be expected on this hardware.
benchmark new smartphones: (Old Thread)
Moderators: hgm, Rebel, chrisw
-
- Posts: 10948
- Joined: Wed Jul 26, 2006 10:21 pm
- Full name: Kai Laskos
Re: benchmark new smartphones:
-
- Posts: 536
- Joined: Thu Mar 09, 2006 12:53 am
Re: benchmark new smartphones:
I downloaded Chess for Android but the user interface is not very intuitive. When I press and hold on the background, a menu pops up but the text is HUGE and I cannot even scroll to display all of the items on it. And it appears I need to somehow download Stockfish 7 separately and then install it?Laskos wrote:This Antutu 6 benchmark seems very good, at least the total score. The SF nps seem very bad, though. Not sure what is this, could you re-run in Chess for Android? If this holds, and there is no another OS issue or bloatware, it would mean "stay away from Snapdragon 820".royb wrote:
OK, I think I found the problem with AnTuTu performance. I had turned on Developer extensions. When I switched them off and re-ran AnTuTu 6.0.1 I got these results:
Score: 134867
Device: SM-G935V
3D: 57356
UX: 40318
CPU: 31106
RAM: 6087
No huge difference though in Droidfish running the SF 7 engine with 4-threads: 1418K nodes per second. I am beginning to think that Droidfish has some issue that's not obvious other than not performing nearly as well as would be expected on this hardware.
I visited Art Bik's webpage and that sent me to the main stockfish webpage where there is only a Stockfish 6 download for Android. This is turning into more effort that I have time for it seems.
It might be time to buy the soon-to-be-announced iPad Pro 9.7" (a.k.a. iPad Air 3 until Apple announces its real name) and just run Smallfish on it.
-
- Posts: 10948
- Joined: Wed Jul 26, 2006 10:21 pm
- Full name: Kai Laskos
Re: benchmark new smartphones:
Here is a fairly normal benchmark I Googled for:royb wrote:
OK, I think I found the problem with AnTuTu performance. I had turned on Developer extensions. When I switched them off and re-ran AnTuTu 6.0.1 I got these results:
Score: 134867
Device: SM-G935V
3D: 57356
UX: 40318
CPU: 31106
RAM: 6087
No huge difference though in Droidfish running the SF 7 engine with 4-threads: 1418K nodes per second. I am beginning to think that Droidfish has some issue that's not obvious other than not performing nearly as well as would be expected on this hardware.
Samsung Exynos 8890 VS Qualcomm Snapdragon 820 in AnTuTu v6.0.1:
The Score - 114 386 VS 131 441
3D - 37589 VS 55198
UX - 42516 VS 38187
CPU - 27833 VS 30594
RAM - 6448 VS 7462
So, your numbers are perfectly reasonable. I don't know what the heck happens with Stockfish, still hope it's an OS or bloatware issue.
-
- Posts: 819
- Joined: Fri Dec 01, 2006 10:46 pm
- Location: Mountain View, CA, USA
- Full name: Aart Bik
Re: benchmark new smartphones:
Well, that is not a non-intuitive interface per-se, but a plain bug! I miss the days I had all possible Android devices on my desk and could test everything worked well everywhere. I will make a fix that scales the dialog better.royb wrote: I downloaded Chess for Android but the user interface is not very intuitive. When I press and hold on the background, a menu pops up but the text is HUGE and I cannot even scroll to display all of the items on it.
Well for obvious reasons I like to keep the GUI separate from any third-party engine. But we try to make your life easier with the Chessbase and open formats that make engines directly discoverable after installing them, see for instance the chess engine package.royb wrote: And it appears I need to somehow download Stockfish 7 separately and then install it?
I visited Art Bik's webpage and that sent me to the main stockfish webpage where there is only a Stockfish 6 download for Android. This is turning into more effort that I have time for it seems.
-
- Posts: 508
- Joined: Fri Jun 04, 2010 7:23 am
Re: benchmark new smartphones:
Qualcomm is simply cheating in the benchmarks. It detects a benchmark and gives full speed (or overclock). When no benchmark is present it underclass or uses standard speed.Laskos wrote:Here is a fairly normal benchmark I Googled for:royb wrote:
OK, I think I found the problem with AnTuTu performance. I had turned on Developer extensions. When I switched them off and re-ran AnTuTu 6.0.1 I got these results:
Score: 134867
Device: SM-G935V
3D: 57356
UX: 40318
CPU: 31106
RAM: 6087
No huge difference though in Droidfish running the SF 7 engine with 4-threads: 1418K nodes per second. I am beginning to think that Droidfish has some issue that's not obvious other than not performing nearly as well as would be expected on this hardware.
Samsung Exynos 8890 VS Qualcomm Snapdragon 820 in AnTuTu v6.0.1:
The Score - 114 386 VS 131 441
3D - 37589 VS 55198
UX - 42516 VS 38187
CPU - 27833 VS 30594
RAM - 6448 VS 7462
So, your numbers are perfectly reasonable. I don't know what the heck happens with Stockfish, still hope it's an OS or bloatware issue.
I don't believe they can make a huge difference in speed that quickly.
This is my opinion anyway. Otherwise, what is the MOST logical explanation?
-
- Posts: 508
- Joined: Fri Jun 04, 2010 7:23 am
Re: benchmark new smartphones:
How did you get 4 cores to work? I can't find an option to set the number of cores. I tried droidfish at the store and s7 is only using 1 core.royb wrote:Julien,
Did you use Droidfish for this result? I have Droidfish with SF 7 as the engine and I get nowhere near what you see. I have the Snapdragon 820-based Galaxy S7 Edge and after one minute with 1 GB of Hash and 4 threads (supposed to be 4 cores on Snapdragon 820), I see this performance in NPS for the starting position: 1096K nodes per second.
I'm wondering if I've got something set up incorrectly with SF 7 that is causing this low a number compared to the result you posted. Yes, your CPU is Exynos 8-core but some have said performance tests have shown that the two are close and the Snapdragon 820 is even a bit faster in some tests.
Roy
Also, maybe try putting the phone in a zip lock bag and then into the freezer (the phone is water proof) for 20-30min. Then try droidfish again. There definitely should be no throttling.
-
- Posts: 508
- Joined: Fri Jun 04, 2010 7:23 am
Re: Budget phone
I think you are referring to g4 or g5 (not g3). The g3 has a snapdragon 801 32bit processor. It gets about 635knps. This was already listed in the chart. This info led someone to mistakingly put 1400 for the g3 in the new chart when that isn't the case.Nordlandia wrote:Smartphone vs Desktop challenge
Question is whether Stockfish 7 on LG G3 will beat Houdini 4 Pro on Pentium E5200.
LG G3 is capable for about 1,400 Knp/s
Pentium E2000 is in the 2,000 range for Stockfish/Komodo and fairly in the 3,000 range for semi-obsolete H4 Pro.
-
- Posts: 536
- Joined: Thu Mar 09, 2006 12:53 am
Re: benchmark new smartphones:
If you touch in the upper left of the screen (near where the word "Player" appears on the GUI), then select menu item "Manage Chess Engines" and the select "Set options", "Threads" is the parameter you can use to change the number of cores used.h1a8 wrote:How did you get 4 cores to work? I can't find an option to set the number of cores. I tried droidfish at the store and s7 is only using 1 core.royb wrote:Julien,
Did you use Droidfish for this result? I have Droidfish with SF 7 as the engine and I get nowhere near what you see. I have the Snapdragon 820-based Galaxy S7 Edge and after one minute with 1 GB of Hash and 4 threads (supposed to be 4 cores on Snapdragon 820), I see this performance in NPS for the starting position: 1096K nodes per second.
I'm wondering if I've got something set up incorrectly with SF 7 that is causing this low a number compared to the result you posted. Yes, your CPU is Exynos 8-core but some have said performance tests have shown that the two are close and the Snapdragon 820 is even a bit faster in some tests.
Roy
Also, maybe try putting the phone in a zip lock bag and then into the freezer (the phone is water proof) for 20-30min. Then try droidfish again. There definitely should be no throttling.
-
- Posts: 508
- Joined: Fri Jun 04, 2010 7:23 am
Re: benchmark new smartphones:
royb wrote:If you touch in the upper left of the screen (near where the word "Player" appears on the GUI), then select menu item "Manage Chess Engines" and the select "Set options", "Threads" is the parameter you can use to change the number of cores used.h1a8 wrote:royb wrote:Julien,
Did you use Droidfish for this result? I have Droidfish with SF 7 as the engine and I get nowhere near what you see. I have the Snapdragon 820-based Galaxy S7 Edge and after one minute with 1 GB of Hash and 4 threads (supposed to be 4 cores on Snapdragon 820), I see this performance in NPS for the starting position: 1096K nodes per second.
I'm wondering if I've got something set up incorrectly with SF 7 that is causing this low a number compared to the result you posted. Yes, your CPU is Exynos 8-core but some have said performance tests have shown that the two are close and the Snapdragon 820 is even a bit faster in some tests.
Roy
How did you get 4 cores to work? I can't find an option to set the number of cores. I tried droidfish at the store and s7 is only using 1 core.
Also, maybe try putting the phone in a zip lock bag and then into the freezer (the phone is water proof) for 20-30min. Then try droidfish again. There definitely should be no throttling.
Thanks! Are you going to try the freezer trick I suggested. I'm curious to know if it's the throttling.
-
- Posts: 196
- Joined: Wed Oct 29, 2014 12:42 am
Re: benchmark new smartphones:
The freezer or the refrigerator trick works okay but not with all processors! In my experiments it worked well with Exynos 4412 running fixed at 1.7ghz but rockchip 3188 can't survive even in a 10 celcius space and freezes at some point. The temperature trick often requires that you lower the gpu voltage and frequency too. Gpu trick helps a lot.
My blog for Android users: http://chesstroid.blogspot.com