DFI ULTRAII-M2 + x2 3600+ Brisbane

Printable View

03-22-2007, 11:34 PM
nzbleach

DFI ULTRAII-M2 + x2 3600+ Brisbane

Ok so here it goes. Basically I had to update the bios of the motherboard to get the cpu to boot at all. Im having the same problems as I did with my old 3800+ x2 cpu. The 667 ram divider does not work so Im having to use the 533 divider and dropping the multi to 9x (default 9.5x) and raising the FSB to get higher ram speed. Now I know this ram does 500mhz on my old CPU when I was running 1:1. Im getting all sorts of funny problems taking this ram above DDR800 using the 533 divider. This is the OC I have managed to get stable so far :
http://i14.photobucket.com/albums/a3...brisbaneoc.jpg

As you can see coretemp is bugging out and not showing the correct temps :

http://i14.photobucket.com/albums/a3...risbaneoc2.jpg

I really want 3G and I hear people with biostar T-force 550 boards talking about MAL ? This is all the options in my bios and if theres anything in there that might be able to improve my OC please let me know. Bare in mind that I've tried 5-5-5 timings for the ram and I've hit a hard wall of 2.85- 2.9ghz. Even the voltage @ 1.55 volts is not getting me past the windows loading BSOD's. I get long beeps when I try 667 divider and when using 533 anything above 2.9ghz is BSOD'ing at the XP loading splash screen.

Genie BIOS Settings:

CPU FSB Ratio Setting............ - x9
CPU n'NB Bus Ratio Setting....... - Auto
CPU Clock Setting................ - 312Mhz
DRAM Speed....................... - DDR 533
PCIE Clock Setting............... - 100Mhz

CPU Voltage Setting.............. - 1.450v
DRAM Voltage Setting............. - 2.2v
Chip Voltage Setting............. - 1.5v
HT Voltage Setting............... - 1.35v

ACPI XSDT Table.................. - Enabled

DRAM Configuration Settings:

1T/2T Memory Timing ............. - 2T
CAS# Latency (Tcl) .............. - 4.0
(Trcd) RAS to CAS R/W Delay ..... - 4 clocks
(Tras) Minimum RAS Active ....... - 4 clocks
(Trc) Row Cycle Time ............ - 12 bus clocks
Trfc for DIMM1 .................. - 16
Trfc for DIMM2 .................. - Auto
Trfc for DIMM3 .................. - Auto
Trfc for DIMM4 .................. - Auto
TwTr Command Delay .............. - Auto
(Twr) Write Recovery Time ....... - Auto
(Trtp) Precharge Time ........... - Auto
(Trrd) RAS to RAS Delay ......... - Auto
DRAM Extra Strength ............. - Level 1
CS/ODT Fine Delay ............... - Auto
CS/ODT Setup Time ............... - Auto
Address/Command Fine Delay ...... - Auto
Address/Command Setup Time ...... - Auto
DRAM Output Driving ............. - Level 3
DQS Training Control ............ - Skip DQS
Memory Hole Remapping ........... - Enabled
Auto Optimize Bottom IO ......... - Enabled

DRAM ECC Enable ................. - Disabled
03-23-2007, 12:24 AM
msimax

try

1T/2T Memory Timing ............. - 2T
CAS# Latency (Tcl) .............. - 4.0
(Trcd) RAS to CAS R/W Delay ..... - 4 clocks
(Tras) Minimum RAS Active ....... - 4 clocks
(Trc) Row Cycle Time ............ - 15 bus clocks
Trfc for DIMM1 .................. - 16
Trfc for DIMM2 .................. - Auto
Trfc for DIMM3 .................. - Auto
Trfc for DIMM4 .................. - Auto
TwTr Command Delay .............. - 3
(Twr) Write Recovery Time ....... - 3
(Trtp) Precharge Time ........... - 3
(Trrd) RAS to RAS Delay ......... - 3
DRAM Extra Strength ............. - Level 3
CS/ODT Fine Delay ............... - Auto
CS/ODT Setup Time ............... - Auto
Address/Command Fine Delay ...... - Auto
Address/Command Setup Time ...... - Auto
DRAM Output Driving ............. - Level 3
DQS Training Control ............ - perform DQS
Memory Hole Remapping ........... - Enabled
Auto Optimize Bottom IO ......... - Enabled
03-23-2007, 01:17 AM
nzbleach

nope

that didnt change anything. weird , 3ghz @ 1.475 volts boots and gets to the windows splash screen but does the neverending XP loading bars. 3G @ 1.5vcore = BSOD. 3G @ 1.450 vcore = no XP load screen, just reboot.
03-23-2007, 02:55 AM
funnyperson1

Try using the 533 divider with 9.5 multi to see if you can get further, I believe the half multi chips don't like being run at even multis and vice versa.
03-23-2007, 04:06 AM
nzbleach

hey mate, looks like your getting similar results. What vcore does it take for you to get to 3g?
03-23-2007, 06:01 AM
MrW

:idea:Just a guess, try lowering HT link. 1.56ghz might be to high.
03-23-2007, 06:08 AM
josty2

Quote:

Originally Posted by MrW

:idea:Just a guess, try lowering HT link. 1.56ghz might be to high.

Quite wierd no-one noticed that yet :P good find! But increasing the Vcore helped didn't it... from no boot, to endless boot to BSOD. Maybe a bit more and it will simply run?? And a good idea is goin to a half multiplier, which has been made for this cpu... might help some...
03-23-2007, 06:11 AM
theelectic

I think there is something funny with AMD CPUs and NForce chipsets with high HTT. With my Asus Crosshair (NForce590) I have used two CPUs, X2 3800+ and Sempron 2800+, and all of them lose multipliers with high HTT, even when the HTT multiplier is drastically lowered. It has something to do with the more obscure RAM timings but I can't figure it out. For example with the Sempron 2800+, if I get it to 300 HTT (300x8) then the DDR2-400, DDR2-533, and DDR2-667 dividers don't work, only the DDR2-800 divider works. Which doesn't make any sense, since both the 667 and 800 dividers are /5! Nothing helps, higher voltage to northbridge, adjusting all kinds of RAM settings, more Vcore, etc.
03-23-2007, 06:12 AM
cumec

Hi!

Very interesting, I have the same board with a x2 4000+ brisbane. I get 2,8GHz with 1,45V vcore stable, but in the region between 2,85 and 2,9Ghz I hit the wall too. 1,55v vcore doesn't change anything...

greetings

cumec
03-23-2007, 06:19 AM
mAJORD

Theelectric.

I've found it fairly progressive between dividers. However I did find an interesting thing with the DFI 590 board, in that there was a HTT range that wouldn't work on one of the dividers. IIRC it was between about 270 and 300.. found by accident.
03-23-2007, 06:27 AM
doompc

Try:

Code:

1T/2T Memory Timing ............. - 2T CAS# Latency (Tcl) .............. - 5.0 (Trcd) RAS to CAS R/W Delay ..... - 5 (Tras) Minimum RAS Active ....... - 15 (Trc) Row Cycle Time ............ - 28 Trfc for DIMM1 .................. - 75ns Trfc for DIMM2 .................. - 75ns Trfc for DIMM3 .................. - 75ns Trfc for DIMM4 .................. - 75ns TwTr Command Delay .............. - 3 (Twr) Write Recovery Time ....... - 6 (Trtp) Precharge Time ........... - 4 (Trrd) RAS to RAS Delay ......... - 3 DRAM Extra Strength ............. - Level 1 CS/ODT Fine Delay ............... - Auto CS/ODT Setup Time ............... - Auto Address/Command Fine Delay ...... - Auto Address/Command Setup Time ...... - Auto DRAM Output Driving ............. - Level 1 DQS Training Control ............ - Skip DQS Memory Hole Remapping ........... - Enabled Auto Optimize Bottom IO ......... - Enabled
03-23-2007, 07:20 AM
theelectic

Quote:

Originally Posted by mAJORD

I've found it fairly progressive between dividers. However I did find an interesting thing with the DFI 590 board, in that there was a HTT range that wouldn't work on one of the dividers. IIRC it was between about 270 and 300.. found by accident.

The weird thing with my NF590 Asus Crosshair is it starts to lose dividers as the HTT is raised. With the Sempron 2800+, up to around 260 HTT all the divders work. Around 270, then the DDR2-400 /8 divider stops working. Around 280, then the DDR2-533 /6 and DDR2-667 /5 go as well. 300+, only the DDR2-800 /5 divider works. Which doesn't make sense since it's the same /5 as DDR2-667. Mind you this is with 2x1GB of RAM, so I was convinced it had something to do with the IMC, but I was told it was something to do with the chipset. With 2x256MB of RAM, all the dividers work and I can get to 370 no problem before my RAM gives out. I experienced something similar when using an X2 3800+ and 2x1GB, as the HTT was raised the dividers would stop working unless I adjusted the clock skew. However with the Sempron NOTHING works.