Gigabyte X48T DQ6 Living Review

[supershanks]

hi Luiman,
try these please.

FSB OverVoltage... 65nm = 1.25v..+: +0.35

try range og +0.15 to 0.25v

CPU Voltage Control...............: 1.39375

try next setting higher.
Play with above & run Prime small ffts stop as soon as 1 core fails. & adjust voltages in bios.

Refresh to ACT Delay (tRFC).......: 68 (seems to stop prior BSOD)

try at 70 and 75

Static tREAD Value (tRD)..........: 9 (seems to stop prior BSOD)

Try at 10 if abover doesn't help.

Run prime Large ffts , see if it runs longer.

Your memory doesn't seem to want to play with the 4.00 memory multiplier. this is a pity as it's only the 4.00 that will let us run your ram at it's rated speed. If above doesn't work, go back to previous settings and try with 3.33D memory multi,

[UNSTABLE_HERO]

I found something for the skews Intel FSB Over Clocking, clock skew and the effects it has on system memory. - OCZ Forum so a higher skew can help without danger!it is not the pll volts as many think and in new gigabyte 45 series theres a seperate pll voltage option and also cpu/nb skews!

[supershanks]

yep pll is present in the bios in the Extreme & P45 boards. Good article I started to use skew at higher FSB, where it helped stability, but with sub 500 fsb I think you can run at 0 skew. In fact driving control which is 800mv by default can be run at 700mv without issues.,

[UNSTABLE_HERO]

Originally Posted by supershanks

yep pll is present in the bios in the Extreme & P45 boards. Good article I started to use skew at higher FSB, where it helped stability, but with sub 500 fsb I think you can run at 0 skew. In fact driving control which is 800mv by default can be run at 700mv without issues.,

no at 3400 with 425fsb only with 200 skews i am stable i tried 0 , 50 , 100 , 150 but the cpu was unstable in prime and occt!my settings are in the forum page 78!its hard to overclock a quad and this board as i can see in many reviews goes perfect with duo cores but not so good with quads!a friend of mine has x48t and we reach 3300 with his q9450 after the bad f5 flash he changed the x48t with asus maximus extreme and now he is stable at 3560 with the same cpu-rams and cpu/nb volts also he has better timmings with his memory than the gigabyte!maybe a good bios can save the day for gigabyte!i see that rev 1.3 reacts very strange in o/c than the rev1.1!!i see that in x38t there are separate bioses for 1.0 rev and 1.3 rev but in x48t one bios for all revs!maybe a new bios can save the day....not like the killer f5 which can f..k the day:lol5 3400 so far i dont want to give more volts so i am gonna try to low the ram timmings and the subtimmings!trd below 9 is a no go!at 1700 ram speed i believe that i could have better timmins with my cellshock but i think i am wrong they have the same performance like my old degraded/bad gskills!now i am running ...Memory Frequency(Mhz).............: 1700
DRAM Timing Selectable ...........: Manual
Standard Timing Control
System Memory Multiplier (SPD)....: 4.00D (4.00B = Bsod at windows)
CAS Latency Time (tCL)............: 8(TRIED 7 AND BUT NO LUCK)
DRAM RAS# to CAS# (tRCD).........: 7
DRAM RAS# Precharge (tRP).........: 7(tried 6 and i see blue screens)
Precharge Delay (tRAS)............: 21
Advanced Timing Control
ACT to ACT Delay(tRRD)............: 5
Rank Write to READ Delay (tWTR)...: 8
Write To Precharge Delay (tWR)....: 12
Refresh to ACT Delay (tRFC).......: 72
Read To Precharge Delay (tRTP)....: 6
Static tREAD Value (tRD)..........: 9(TRIED 7 AND 8 BUT NO LUCK)
Static tREAD Phase Adj............: Auto
Command Rate (CMD)................: 2T
MCH+0.250
RAM VOLT+0.35

[supershanks]

Yes your probably correct, the skew helps quad stability, but isn't needed till higher 540+ fsb on duo cpu.
Another thing looking at both yourself & Luiman, running quads, your running lower FSB, which in turn means that to maximise your ddr3 speeds you need to use the 4.00x memory multiplier.
This 4.00 Multiplier doesn't appear to run as well as the 3.33 which i've run almost exclusively.

does Static tREAD Value (tRD) =10 run ?

[UNSTABLE_HERO]

yep Static tREAD Value (tRD) =10 runs fine!iam stable at occt(4 hours)prime(12 hours)but in games(15 minutes)i get bsob !i think its the pcie volts i must give +0.20 because when i was running prime/occt i had +0.10 the pcie once i tried normal i could not pass any test now in game th vga works 100% maybe pcie needs +0.20v @ 100mhz or its something else?i hope not!

[supershanks]

if it's that stable at at trd=10, then upping pci-e volts won't harm , what PCI-E Frequency are you running ? Auto 100 or ?? I generally run 103.

getting bsod like that , I wonder if it might be a gfx driver issue or gfx heat ??

[UNSTABLE_HERO]

CPU Clock Ratio...................: 8x
Fine CPU Clock Ratio...........: +0.0
CPU Frequency.....................: 3.4Ghz
CPU Host Clock Control............: Enabled
CPU Host Frequency(Mhz)...........: 425
PCI Express Frequency (Mhz).......: 104
C.I.A.2...........................: Disabled
Performance Enhance...............: Standard
System Memory Multiplier (SPD)....: 4.00D
Memory Frequency(Mhz).............: 1700
DRAM Timing Selectable ...........: Manual
Standard Timing Control
CAS Latency Time (tCL)............: 8
DRAM RAS# to CAS# (tRCD).........: 7
DRAM RAS# Precharge (tRP).........: 7
Precharge Delay (tRAS)............: 21
Advanced Timing Control
ACT to ACT Delay(tRRD)............: 5
Rank Write to READ Delay (tWTR)...: 8
Write To Precharge Delay (tWR)....: 12
Refresh to ACT Delay (tRFC).......: 72
Read To Precharge Delay (tRTP)....: 6
Static tREAD Value (tRD)..........: 9
Static tREAD Phase Adj............: Auto
Command Rate (CMD)................: 2T
Clock Driving and Skew Control
CPU/PCIEX Clock Driving Control...: 800mv
CPU Clock Skew Control............: 200PS
(G)MCH Clock Skew Control.........: 200PS
CPU GTLREF1 Voltage*.......(67%)..: auto
CPU GTLREF2 Voltage*.......(61%)..: Normal
System Voltage Control () = Base Voltages
DDR3 OverVoltage Control...(1.50v): +0.35
PCI-E OverVoltage Control.,(1.50v): +0.10
FSB OverVoltage.......... = 1.15v.: +0.20
(G)MCH OverVoltage Control,(1.25v): =0.250
Loadline Calibration..............: Enable
CPU Voltage Control...............: +1.29300v
....Advanced Bios Features
Limit CPUID Max to 3..............: Disable
No-Execute Memory Protect.........: ENABLE
CPU Enhanced Halt(CIE)............: Disable
CPU Thermal Monitor................: Disable
CPU EIST Function.................: Disable
Virtuaalization Technology........: Disable

THESE ARE MY STABLE SETTINGS!everybody with this m/b listen what happened to me!i was running stable after hours of testing i was soooooo stable!i say hey lets try to tight the timmings i put tdr at 8(BSOB) i try 1t(stable)i try tRP at 6(BSOB) I GO BACK TO BIOS AND PUT EVERYTHING BACK TO MY STABLE SETTNIGS AFTER THAT I HAD GAME CRASHES AND BLUE SCREENS I TRIED TO RUN OCCT AND FAILED AFTER 30 SECS!I SAY HEY I AM GOING MAD WITH THIS M/B THESE WAS MY STABLE SETTINGS WHAT THE F..K IS GOING ON NOW!i go to bios and i load optimized defaults in bios,after that i put again my stable settings and now i am stable no more bsobs!the lesson is never have thrust buggy gigabytes bioses!

[supershanks]

End of rant ?
Dropping tRD to 8 will probably have corrupted cmos, due to memory being unstable So when you recover & try your old stable timings, they don't work. Best solution is to clear cmos & re enter or reload settings.
Can happen on any mobo.

[UNSTABLE_HERO]

its not a rant but just a warning that this mb is so crazy i have a pc from 1889 my first cpu was a 8088!at 1998 i starded to o/c(sometimes and vmod) my systems till 2006!frOm 2006 to 2008 i lose contact with the cpus,vgas,mbs and o/c because i was at the army!i have changed many motherboards before and everything was so clear at o/c(cpu timmings rams volts BUT NOT CRAZY corrupts at cmos from bios) but this is the worse mb ever(i dont want to tell again about the bad f5 bios which killed my first mb)of course this corrupts can happen to any mobo by 1 out of 10 times but at my x48t happens 9 out of 10 times!i see my friends with other mobos with same cpu model and ram to run so stable at higher speeds when i helped them to o/c and i see a nice stable bios/mb!a friend had an x48t and we reached 3300 with his q9450 after the bad f5 flash he changed the x48t with asus maximus extreme and now he is stable at 3560 with the same cpu-rams and little less cpu/nb volts also he has better timmings with his memory on his new mobo!i also o/ced a 9550 very easy on a msi x48 platinium!i was expecting more from this expensive mb!and the bios should have votlage settings of p45 series they dont have pll volts(and many more useful settings)which was very useful to my o/c tests on the asus and msi boards!so once again thanks to supershanks for his useful help and this great forum