21 Replies Latest reply on Jun 25, 2015 9:12 AM by exceed

    Any ideas for fixing Quadfire (CrossfireX) on my build?

    exceed

      Currently I've just upgraded my computer to the following specification:

       

      - Intel i7 5960X.

      - 32GB DDR4 RAM (XMP set at 3Ghz).

      - Asrock X99 Extreme 11 Motherboard (there are two 4 pin molex pins that are being supplied power from PSU - required for 4 graphics cards). Motherboard is also on the latest 1.20 bios version.

      - 2 x 295X2's in Crossfire (1st card in 1st slot, 2nd card in 4th slot - as per the manual - both running at 16X PCIE rate).

      - EVGA G2 1600W (both cards are connected to the GPU power ports with 2 x 8 pin PCIE).

      - LG 34UC97-S 34" x 2 monitors (both displays connected to the card in 1st slot).

       

      For the life of me I can't get Quadfire to work, whenever I launch any games

       

      I've run DDU (display driver uninstaller) several times and installed catalyst 15.3, 15.4 and 15.5 with no difference.

       

      For some reason, benchmarks work okay:

      - Firestrike: AMD Radeon R9 295X2 video card benchmark result - Intel Core i7-5960X,ASRock X99 Extreme11

      - Firestrike Extreme: AMD Radeon R9 295X2 video card benchmark result - Intel Core i7-5960X,ASRock X99 Extreme11 

      - Firestrike Ultra: AMD Radeon R9 295X2 video card benchmark result - Intel Core i7-5960X,ASRock X99 Extreme11

       

      In CCC, I've:

      - enabled CrossfireX.

      - disabled vsync.

       

      I've also turned off ULPS mode which allows me to see the temperature of all cores (which previously showed as 0). But usage for GPU 1 and 2 just do not move in any games, whilst 3 and 4 fluctuate (as expected).

       

      And ensured that I'm running full screen in all games.

       

      The slave 295X2 heatsink/watercooling radiator doesn't get hot at all, whilst the master gets very hot (again as expected) so it's not as if the utilisation graphs are incorrect.

       

      I'm yet to view MSI afterburner GPU usage whilst running 3dmark to verify any difference.

       

      Anyone have any ideas?

       

      Thanks in advance...