eva2000
07-14-2008, 07:11 AM
Playing with beta bioses for DFI LP UT X48-T3RS and Sascha would be pleased to know the Cellshock PC3-14400 are now working :)
Using DDRAM Default Skew Model 0 (default)
http://fileshosts.com/intel/DFI/DFI_UT_X48_T3RS/results/Nautilus500/E8500_Q740A493T_1808/cellshock/14400/8x/400_1333/1t/AFD_500-833_65415-A-52-pl9_eaeee_12-12_odt1_3_allauto_1.3375-2-1.655-1.35_model0/spi32m_11m37s954ms.png
click image for full screenshot
http://fileshosts.com/intel/DFI/DFI_UT_X48_T3RS/results/Nautilus500/E8500_Q740A493T_1808/cellshock/14400/8x/400_1333/1t/AFD_500-833_65415-A-52-pl9_eaeee_12-12_odt1_3_allauto_1.3375-2-1.655-1.35_model0/hd3850cf/3dmark06_810_1000_18333_tn.png (http://fileshosts.com/intel/DFI/DFI_UT_X48_T3RS/results/Nautilus500/E8500_Q740A493T_1808/cellshock/14400/8x/400_1333/1t/AFD_500-833_65415-A-52-pl9_eaeee_12-12_odt1_3_allauto_1.3375-2-1.655-1.35_model0/hd3850cf/3dmark06_810_1000_18333.png)
Weird part is memtest86+ v2.01 reports 400/1333 divider as 400/1200 for 750mhz speeds and not 833mhz, and super pi 32m seems slow for those clocks by 15-25s taking into account loose tRD PL.
Hopefully will get to bottom of this :)
Using DDRAM Default Skew Model 0 (default)
http://fileshosts.com/intel/DFI/DFI_UT_X48_T3RS/results/Nautilus500/E8500_Q740A493T_1808/cellshock/14400/8x/400_1333/1t/AFD_500-833_65415-A-52-pl9_eaeee_12-12_odt1_3_allauto_1.3375-2-1.655-1.35_model0/spi32m_11m37s954ms.png
click image for full screenshot
http://fileshosts.com/intel/DFI/DFI_UT_X48_T3RS/results/Nautilus500/E8500_Q740A493T_1808/cellshock/14400/8x/400_1333/1t/AFD_500-833_65415-A-52-pl9_eaeee_12-12_odt1_3_allauto_1.3375-2-1.655-1.35_model0/hd3850cf/3dmark06_810_1000_18333_tn.png (http://fileshosts.com/intel/DFI/DFI_UT_X48_T3RS/results/Nautilus500/E8500_Q740A493T_1808/cellshock/14400/8x/400_1333/1t/AFD_500-833_65415-A-52-pl9_eaeee_12-12_odt1_3_allauto_1.3375-2-1.655-1.35_model0/hd3850cf/3dmark06_810_1000_18333.png)
Weird part is memtest86+ v2.01 reports 400/1333 divider as 400/1200 for 750mhz speeds and not 833mhz, and super pi 32m seems slow for those clocks by 15-25s taking into account loose tRD PL.
Hopefully will get to bottom of this :)