• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Rat GPU Benchmark

4 x 290Xs @1250/1625

4930k @4.8

Low
SQcVzP2.jpg


Medium
BR6O95f.jpg


High
fwGjsX2.jpg
 
High preset
4770K @4.6
REF 290X @1150/1500.

U0ZH5qY.jpg.png

Medium preset
Low preset

Looking at the scores currently for single cards this seem wrong:confused:

Anything else i need to do other than just hit the high preset?:confused:

I think you need to untick the Hawaii HT box before you run the bench. I think this maybe the integrated graphics on the 4770k.

With X79 this is no Hawaii HT box to tick.:)
 
Don't think it's that Kaaps, take a look.
RVVDGv8.jpg.png
Disabling that looks like it's something to do with my card, the other boxes seem to control the IGPU usage:confused:

Will the bench not run just ticking the top box on it's own ?

This may be a Haswell thing as what springs to mind is Haswell CPUs allow you to do extra things in RTW2.

What do you see as the bench is running, is it filling 1 or 2 squares are it runs. On mine with quadfire I see it filling 4 squares as it runs.
 
Last edited:
It runs kaaps, but if i untick that HT box the times are terrible (around 68 seconds on low).

I think it must be a Haswell thing, somehow the card is making use of the idle HT on the CPU or something.

I am going to put your scores on the scoreboard as it looks a bit like using X79 on the futuremark benches where you get a huge score compared to a normal quad because of the physics the hex cores give you.
 
I notice that some of the earlier entries are using v0.6.0 of the bench while Avenged7Fold and uksoldierboy are using v0.7.0 Beta ? Can't click the link in work but it might be worth having a gander at the change log :confused:

+1

Looks like the version to use is v0.70

v0.7.0 Beta

Added 3dsmax / Maya 2015 support.

Fixed a problem that was causing to render with less GPUs than it should be when the number of video cards was greater than the CPU cores.

Now we use two active tiles per OpenCL device to maximize the GPU usage ( you'll notice a 'HT' signature on the device list to reflect this, aka GPU 'HyperThreading' :D ).

Recompiled using the latest libraries and compiler ( VS2013 Update 2, Qt 5.3.0, tbb42_20140416oss )
 
Anybody tried the v0.7.1 Beta yet? Not sure what to run now :rolleyes:

I read in the notes that it can run slightly slower than v0.7.0

v0.7.1 Beta

Optimized a 3% the rendering speed for AMD cards and Catalyst 14.4.

Added the app's version text to the benchmark final result dialog.

Increased a bit some maths operations's precision because we were getting some problems with reflections. This may slow down a bit the rendering speed but it was really needed.
 
Back
Top Bottom