Results 11 to 16 of 16
Thread: OpenCL GPU&CPU?
-
04-07-2014 06:48 PM #11
Nodens PC Specs Laptop (Model) ASUS GL702VW-GC004T Motherboard Rampage V Extreme Edition 10 Processor Intel 3930K C2 Memory (part number) CMZ16GX3M4X2133C11 Graphics Card #1 ASUS STRIX-GTX970-DC2OC-4GD5 Monitor ASUS VG278H Storage #1 3x OCZ Vertex 3 240 (RAID5 - LSI 9361-8i) Storage #2 3x WD2003FYYS (RAID5 - LSI 9361-8i) CPU Cooler Corsair H100i + 4x Noctua NF-P12 Case Coolermaster HAF-X Power Supply Seasonic X-1250 Keyboard Razer Anansi Mouse Logitech G502 OS Windows 10 Pro x64 Network Router Linksys E4200v2 (running TomatoUSB/Toastman firmware)
- Join Date
- Mar 2012
- Reputation
- 266
- Posts
- 4,389
RAMPAGE Windows 8/7 UEFI Installation Guide - Patched OROM for TRIM in RAID - Patched UEFI GOP Updater Tool - ASUS OEM License Restorer
There are 10 types of people in the world. Those who understand binary and those who don't!
RealBench Developer.
-
04-09-2014 11:10 PM #12
- Join Date
- Oct 2013
- Reputation
- 10
- Posts
- 25
There is something else going on besides the driver. I think the wrong switch may be used because I can run my CPU OpenCL standalone but it doesn't automatically get checked when running the "realbench" version. I get around 1060 score in luxmark with all OpenCL GPU's + CPU's enabled so fixing this will greatly increase my score.
Not that I already have the most powerful G46VW here or anything already
EDIT: Here is a comparison of the Luxmark scores..
540 ish for "RealBench" version and 1060 ish Running manually
Last edited by ha1o2surfer; 04-11-2014 at 03:04 PM.
G46VW (Now with THUNDERBOLT!!), 3840QM (overclocked, cTDP, 72w), Modded BIOS, 16GB, Samsung 750GB SSD, 1TB (Samsung mSATA) Liquid metal thermal paste
-
04-11-2014 04:22 PM #13
- Join Date
- Oct 2013
- Reputation
- 10
- Posts
- 25
Not to be a pest but I also noticed one more thing. While running the real bench version of luxmark I got 540ish as you konw from above. But If I run the same settings on standalone luxmark (all OpenCL GPU's) I get around 675. Why such lower numbers in the real bench version of luxmark? I would suggest letting luxmark use all devices as it can test throttling (if any)
G46VW (Now with THUNDERBOLT!!), 3840QM (overclocked, cTDP, 72w), Modded BIOS, 16GB, Samsung 750GB SSD, 1TB (Samsung mSATA) Liquid metal thermal paste
-
05-06-2014 02:33 PM #14
Nodens PC Specs Laptop (Model) ASUS GL702VW-GC004T Motherboard Rampage V Extreme Edition 10 Processor Intel 3930K C2 Memory (part number) CMZ16GX3M4X2133C11 Graphics Card #1 ASUS STRIX-GTX970-DC2OC-4GD5 Monitor ASUS VG278H Storage #1 3x OCZ Vertex 3 240 (RAID5 - LSI 9361-8i) Storage #2 3x WD2003FYYS (RAID5 - LSI 9361-8i) CPU Cooler Corsair H100i + 4x Noctua NF-P12 Case Coolermaster HAF-X Power Supply Seasonic X-1250 Keyboard Razer Anansi Mouse Logitech G502 OS Windows 10 Pro x64 Network Router Linksys E4200v2 (running TomatoUSB/Toastman firmware)
- Join Date
- Mar 2012
- Reputation
- 266
- Posts
- 4,389
Any comparison with the original Luxmark should not be made. For everal reasons:
1) The versions are different (you're comparing 2.0 vs 2.1 which is not available as a binary but only if you compile it from git source.
2) RB version is different from 2.1 git version as well as it depends on Qt5, instead of Qt4 and it and its dependencies are built with different toolchains. As in RB's version is modified.
3) The scene we are rendering is different as well, even if 1 and 2 were not the case, the results would still be different.
Bottomline is do not compare the OpenCL test results of RB with plain Luxmark. Period.RAMPAGE Windows 8/7 UEFI Installation Guide - Patched OROM for TRIM in RAID - Patched UEFI GOP Updater Tool - ASUS OEM License Restorer
There are 10 types of people in the world. Those who understand binary and those who don't!
RealBench Developer.
-
06-10-2014 10:55 PM #15
- Join Date
- Oct 2013
- Reputation
- 10
- Posts
- 25
I see what you are saying but even when it comes to the test, RB claims my CPU is at 100% where taskmanager says 15-25% It's only using the Intel HD 4000 and the GTX 660m where as I thought it would also use the CPU as well. That is where my confusion lies.
G46VW (Now with THUNDERBOLT!!), 3840QM (overclocked, cTDP, 72w), Modded BIOS, 16GB, Samsung 750GB SSD, 1TB (Samsung mSATA) Liquid metal thermal paste
-
06-30-2014 04:56 PM #16
- Join Date
- Nov 2013
- Reputation
- 10
- Posts
- 2
I'm new around here. So forgive me this question but i must ask. I'm not really understood earlier post or on the platform i5 3570k Maximus V Extreme and GTX 770 GPU should i have switch on graphic intel HD 4000 to raise my score in RealBench? I currently have turned off and uninstalled the drivers intel hd graphics using the tools display driver uninstaller. Having 2 graphics enabled make me some errors and I had just turned off the built-in CPU. Thank you very much for your answer. And to how enable in OpenCL part of benchmark CPU and GPU ....
Last edited by tzp0slk; 06-30-2014 at 09:53 PM.
What do you fear most... in the world?
The possibility that love is not enough.