DFI final S939 Motherboard CFX3200 ATI RD580 results

[LivingL@rGe81]

I am having trouble getting 250Mhz 2-2-2-5 stable on any of my UTT BH-5. I have gskill and ocz 3200 gold and I keep getting windows memtest errors and memtest 1.65 errors. I have tried DS @ DDS of 1-4 for both and nothing seems to help. I am running my sata on the ULI controller though so maybe it is what Tony is talking about. I did manage to clear alot of weird things up by resetting my cmos using killcmos and flashing 0417 bios 3 times in a row but my memory clocking is still kinda low at 230Mhz 2-2-2-5. Im sure its all in the settings, anyone have any suggestions?

I've found the best settings so far i've found are -

DS=3
DDS=2


Hope this helps

[babalouj]

@Living: Hey man, thanks so much for your help! I will try those setting when I get home from work. I have a couple of questions still though. What TREF are you using, I dont know what 166Mhz, 1.95us corresponds to in the bios. Also, what slots are you using, orange or yellow?

@Everyone else: Can anyone give me some settings for 4x512 UTT bh-5, I am thinking of running my gskill and ocz together. I know it will run 2T but I think the benefits of 2gb will out weight 2T.

[LivingL@rGe81]

@Living: Hey man, thanks so much for your help! I will try those setting when I get home from work. I have a couple of questions still though. What TREF are you using, I dont know what 166Mhz, 1.95us corresponds to in the bios. Also, what slots are you using, orange or yellow?

I'm using the orange slots but haven't tested the yellow ones yet!

TREF is 4708 in bios

[spajdr]

Odd.

No. Oblivion is great here. No issues. 6.3 cats and the "chuck patch" as of late. 4x AA + HDR and 16x HQ AF. So far I am loving Xfire. I rename the .exe AFR-FriendlyD3D.exe and click on it to start the game. That enables AFR Xfire support. I picked the tip up from the FiringSquad>

Yeah i used that name of .exe too .. doesnt matter what catalysts i used, i seen flashy ground with that name. Most probably faulty card.

If i used fear or sam2.exe it was ok but then inventory or pause makes artefacts in background.

[babalouj]

@Living: Your settings worked great! I can run windows memtest and memtest 1.65 at 250Mhz 2-2-2-5 just fine now but my comp randomly restarts now, my guess is that its a drive strength thing.

Anyone have any 4x512 BH-5 utt settings for this board?

[spajdr]

still messing with settings, but so far i love this mobo, third love in row .. err third DFI mobo that is

[eva2000]

Just started playing with board and BH-5/UTT BH-5 ram - anyone notice even at 3.4v vdimm ram runs alot hotter on this board than DFI NF4 Ultra-D/SLI-D/SLI-DR series ? I'm talking like TCCD at 3.0v vdimm hot!

[LivingL@rGe81]

Just started playing with board and BH-5/UTT BH-5 ram - anyone notice even at 3.4v vdimm ram runs alot hotter on this board than DFI NF4 Ultra-D/SLI-D/SLI-DR series ? I'm talking like TCCD at 3.0v vdimm hot!

I noticed the same thing! Its almost like its pushing more current throught the ram???

[bazx]

Just started playing with board and BH-5/UTT BH-5 ram - anyone notice even at 3.4v vdimm ram runs alot hotter on this board than DFI NF4 Ultra-D/SLI-D/SLI-DR series ? I'm talking like TCCD at 3.0v vdimm hot!

cant say i have but i have noticed that bh5 oc better on the m/b than my dr

[eva2000]

2x 512MB OCZ PC3200 Gold UTT BH-5 @255mhz 2-2-2-5 7-14-2223 at 3.4v vdimm

Last night tested without the dual lcd thermal probes and only 2x40mm fans and modules would be scorching hot and test #5 would fail at 2nd to 4th loop with a handful of errors.

Decided to put my dual lcd thermo probe to work and attached it to the side of each of my 2x 512MB OCZ PC3200 Gold UTT BH-5 and use a thermaltake 90+ cfm 120x25mm fan to blow directly on the modules instead of my 2x 40mm fans.

Memtest86+ v1.65
5 loops of test #5 and temps already peak at
Dimm 1 module = 37.5C
Dimm 3 module = 32C

So far up to 90+ loops and no errors unlike last night!

Room temp between = 27.5C to 28.8C open desktop benching system setup.

Conclusion, Dimm 1 slot is a hot bugger and that with even 3.4v vdimm you better have at least 90cfm of fan cooling on your modules to keep memory modules' temps in check for stability!

[eva2000]

I've only had my CFX3200-DR for about ~24hrs but so far results look promising http://i4memory.com/showthread.php?t=2629 with alot of tips and tweaks from Tony at OCZ it has helped speed up the learning curve - which is very steep due to the amount of bios settings to tweak!

i.e. 2x 512MB Corsair XMS3500C2v1.1 BH-5

Click on image for the larger full picture.

==============================
10x300HTT = 3000mhz
@250mhz 2-2-2-5 7-14-2222 at 3.44v

FSB Bus Frequency: 300
LDT/FSB Frequency Ratio: 4x (tried 5x but system didn't boot even with NMOS/PMOS drive strength @31 so dropped it to 4x)
CPU/FSB Frequency Ratio: 10x
RD580 HT PLL Speed: High
HT Bus NMOS Drive Strength: 5
HT Bus PMOS Drive Strength: 5
HT Bus Receiver Impendence: 5
CPU HT Bus Drive Strength: Strong
K8 Cool & Quiet Support: Disabled

CPU VID Control: 1.5v (set higher due to low vcore + high vdimm issue)
CPU VID Special Control: AUTO
DRAM Voltage Control: 3.44v
SB PCIE Voltage: 1.8v
NB Analogue Voltage: 1.34v
LDT Bus Voltage: 1.34v
NB Core Voltage: 1.34v

DRAM Frequency Set: 166
CPC: Enabled
CAS (tCL): 2
tRCP: 2
tRAS: 5
tRP: 2
tRC: 7
tRFC: 14
tRRD: 2
tRW: 2
tRTW: 2
tWTR: 2
tREF: 4708
Bank Interleaved: Enabled
Errata 94: Disabled
Errata 123: AUTO
Odd Divisor Correct: Disabled

DQS Skew Control: AUTO
DQS Skew Value: N/A
DRAM Drive Strength: 3 (weak)
DRAM Data Drive Strength: 2 (-33%)
DIMM 1/2 Clock Timing Skew: AUTO
DIMM 3/4 Clock Timing Skew: 300ps delay
Max Async Latency: 8ns
Read Preamble Time: 5ns
IdleCycle Limit: 16clks
Dynamic Counter: Disabled
R/W Queue Bypass: 16x
Bypass Max: 7x
Burst Length: 4

Super Pi v1.5 32M



==============================
10x300HTT = 3000mhz
@250mhz 2-2-2-5 7-14-2222 at 3.44v

FSB Bus Frequency: 300
LDT/FSB Frequency Ratio: 4x (tried 5x but system didn't boot even with NMOS/PMOS drive strength @31 so dropped it to 4x)
CPU/FSB Frequency Ratio: 10x
RD580 HT PLL Speed: High
HT Bus NMOS Drive Strength: 5
HT Bus PMOS Drive Strength: 5
HT Bus Receiver Impendence: 5
CPU HT Bus Drive Strength: Strong
K8 Cool & Quiet Support: Disabled

CPU VID Control: 1.5v (set higher due to low vcore + high vdimm issue)
CPU VID Special Control: AUTO
DRAM Voltage Control: 3.44v
SB PCIE Voltage: 1.8v
NB Analogue Voltage: 1.34v
LDT Bus Voltage: 1.34v
NB Core Voltage: 1.34v

DRAM Frequency Set: 166
CPC: Enabled
CAS (tCL): 2
tRCP: 2
tRAS: 5
tRP: 2
tRC: 7
tRFC: 14
tRRD: 2
tRW: 2
tRTW: 2
tWTR: 2
tREF: 4708
Bank Interleaved: Enabled
Errata 94: Disabled
Errata 123: AUTO
Odd Divisor Correct: Disabled

DQS Skew Control: AUTO
DQS Skew Value: N/A
DRAM Drive Strength: 3 (weak)
DRAM Data Drive Strength: 2 (-33%)
DIMM 1/2 Clock Timing Skew: AUTO
DIMM 3/4 Clock Timing Skew: 300ps delay
Max Async Latency: 8ns
Read Preamble Time: 5ns
IdleCycle Limit: 16clks
Dynamic Counter: Enabled
R/W Queue Bypass: 16x
Bypass Max: 7x
Burst Length: 4

Super Pi v1.5 32M



==============================
10x300HTT = 3000mhz
@250mhz 2-2-2-5 7-12-2212 at 3.44v

FSB Bus Frequency: 300
LDT/FSB Frequency Ratio: 4x (tried 5x but system didn't boot even with NMOS/PMOS drive strength @31 so dropped it to 4x)
CPU/FSB Frequency Ratio: 10x
RD580 HT PLL Speed: High
HT Bus NMOS Drive Strength: 5
HT Bus PMOS Drive Strength: 5
HT Bus Receiver Impendence: 5
CPU HT Bus Drive Strength: Strong
K8 Cool & Quiet Support: Disabled

CPU VID Control: 1.5v (set higher due to low vcore + high vdimm issue)
CPU VID Special Control: AUTO
DRAM Voltage Control: 3.44v
SB PCIE Voltage: 1.8v
NB Analogue Voltage: 1.34v
LDT Bus Voltage: 1.34v
NB Core Voltage: 1.34v

DRAM Frequency Set: 166
CPC: Enabled
CAS (tCL): 2
tRCP: 2
tRAS: 5
tRP: 2
tRC: 7
tRFC: 12
tRRD: 2
tRW: 2
tRTW: 1
tWTR: 2
tREF: 4708
Bank Interleaved: Enabled
Errata 94: Disabled
Errata 123: AUTO
Odd Divisor Correct: Disabled

DQS Skew Control: AUTO
DQS Skew Value: N/A
DRAM Drive Strength: 3 (weak)
DRAM Data Drive Strength: 2 (-33%)
DIMM 1/2 Clock Timing Skew: AUTO
DIMM 3/4 Clock Timing Skew: 300ps delay
Max Async Latency: 8ns
Read Preamble Time: 5ns
IdleCycle Limit: 16clks
Dynamic Counter: Enabled
R/W Queue Bypass: 16x
Bypass Max: 7x
Burst Length: 8

Super Pi v1.5 32M



==============================
10x300HTT = 3000mhz
@250mhz 2-2-2-0 7-12-0211 at 3.44v

Memtest86+ v1.65 bandwidth = 3006MB/s

FSB Bus Frequency: 300
LDT/FSB Frequency Ratio: 4x (tried 5x but system didn't boot even with NMOS/PMOS drive strength @31 so dropped it to 4x)
CPU/FSB Frequency Ratio: 10x
RD580 HT PLL Speed: High
HT Bus NMOS Drive Strength: 5
HT Bus PMOS Drive Strength: 5
HT Bus Receiver Impendence: 5
CPU HT Bus Drive Strength: Strong
K8 Cool & Quiet Support: Disabled

CPU VID Control: 1.5v (set higher due to low vcore + high vdimm issue)
CPU VID Special Control: AUTO
DRAM Voltage Control: 3.44v
SB PCIE Voltage: 1.8v
NB Analogue Voltage: 1.34v
LDT Bus Voltage: 1.34v
NB Core Voltage: 1.34v

DRAM Frequency Set: 166
CPC: Enabled
CAS (tCL): 2
tRCP: 2
tRAS: 0
tRP: 2
tRC: 7
tRFC: 12
tRRD: 0
tRW: 2
tRTW: 1
tWTR: 1
tREF: 4708
Bank Interleaved: Enabled
Errata 94: Disabled
Errata 123: AUTO
Odd Divisor Correct: Disabled

DQS Skew Control: AUTO
DQS Skew Value: N/A
DRAM Drive Strength: 3 (weak)
DRAM Data Drive Strength: 2 (-33%)
DIMM 1/2 Clock Timing Skew: AUTO
DIMM 3/4 Clock Timing Skew: 150ps delay
Max Async Latency: 8ns
Read Preamble Time: 4.5ns
IdleCycle Limit: 0clks
Dynamic Counter: Enabled
R/W Queue Bypass: 16x
Bypass Max: 7x
Burst Length: 8

Super Pi v1.5 32M


The above 23min 59.297s time managed to beat my DFI NF4 Ultra-D time of 23min 59.641s.



Everest Home:
Read = 7,142 MB/s
Write = 3,361 MB/s
Latency = 36.2 ns

Sandra 2005 Lite SR3
Buffered = 7,486 / 7,419
Unbuffered = 3,717 / 3,847


==============================

working on figuring out 10x315HTT right now with dividers - anyone got any tips for >300HTT on dividers ?

[texuspete00]

My OCs are running shady on a known good OC'ing 170. Wish I could find someone using Micron 5bD's anymore. I know, patience and tweaking. Guess a good OC will have to wait for the weekend.

[eva2000]

Guys any tips on high HTT stability ?

High HTT/FSB on divider tests.

Seems so far managed 366FSB/HTT as the highest i've been able to boot into windows. It needed LDT dropped from 4x to 3x though. Managed to push out an 8M time but windows frooze trying to save the cpuz validation file heh.

But seems 334FSB/HTT is fine with LDT 4x!. Haven't tried higher than 334FSB/HTT for now.

Click on image for large full size screenie with A64Info showing the HTT Bus speed etc



Settings used for 9x334HTT = 3005Mhz

FSB Bus Frequency: 334
LDT/FSB Frequency Ratio: 4x (tried 5x but system didn't boot)
CPU/FSB Frequency Ratio: 9x
RD580 HT PLL Speed: High
HT Bus NMOS Drive Strength: 31
HT Bus PMOS Drive Strength: 31
HT Bus Receiver Impendence: 10
CPU HT Bus Drive Strength: Strong
K8 Cool & Quiet Support: Disabled

CPU VID Control: 1.5v (set higher due to low vcore + high vdimm issue)
CPU VID Special Control: AUTO
DRAM Voltage Control: 3.4v
SB PCIE Voltage: 1.85v
NB Analogue Voltage: 1.34v
LDT Bus Voltage: 1.34v
NB Core Voltage: 1.34v

DRAM Frequency Set: 150
CPC: Enabled
CAS (tCL): 2
tRCP: 2
tRAS: 5
tRP: 2
tRC: 7
tRFC: 14
tRRD: 2
tRW: 2
tRTW: 2
tWTR: 2
tREF: 4708
Bank Interleaved: Enabled
Errata 94: Disabled
Errata 123: AUTO
Odd Divisor Correct: Disabled

DQS Skew Control: AUTO
DQS Skew Value: N/A
DRAM Drive Strength: 3 (weak)
DRAM Data Drive Strength: 2 (-33%)
DIMM 1/2 Clock Timing Skew: AUTO
DIMM 3/4 Clock Timing Skew: 450ps
Max Async Latency: 9ns - needed 9ns otherwise memtest86+ test #5 errors with 8ns
Read Preamble Time: 5ns
IdleCycle Limit: 16clks
Dynamic Counter: Enabled
R/W Queue Bypass: 16x
Bypass Max: 7x
Burst Length: 4

[spajdr]

hmm thats weird ... well i only have FSB 300 but with LDT 5x .. so 1500Mhz and its rock stable.

[spajdr]

guys i didnt reinstalled XP since i got DFI (before i got asus xpress 3200) but exactly once a hour, my PC froze, not entirely, but sound is repeating for like 10-15seconds then its fine for another hour, where could be the problem? it happened with asus too, but not so long and frequent.

[eva2000]

tried dropping to LDT 4x ?

Anyone finding CFX3200-DR needs more vcore to the cpu to get same prime95/stressprime 2004 stable clocks than DFI NF4 series ?

I'm needing 0.1v more vcore windows on CFX3200-DR than on DFI NF4 for my Opteron 146 which did 3000mhz at 1.42-1.47v on DFI NF4 Ultra-D now needs 1.52v on CFX3200-DR ?

Same Opteron 146 CAB2E 0540FPMW on same water cooling setup = Corsair Hydrocool 200 with s478 to s939 adaptor which gives very consistent and repeatable mounting pressure every mount as you can see from below pic.



DFI NF4 Ultra-D R.AD0 = 3150mhz max 32M at 1.616v
DFI CFX3200-DR R.A00 = 3100mhz max 32M at 1.616v

Wonder if it can come down to memory settings on this board ? Might try a remount of the waterblock.

[El Snorro]

@Eva2000

No, my 3700+ 0505 and my 4000+ 0505 are both running fine at 3ghz 1,39v under TTBT.

Just like in Nf4 ultra-d

[situman]

Should I or should I not? I'm certainly no expert when it comes to tuning. It does seem like this is the best board for 939. So far no fried this burned that flames all over. It would be the last 939 anyone would need or want to play with.

[Raybo]

Should I or should I not? I'm certainly no expert when it comes to tuning. It does seem like this is the best board for 939. So far no fried this burned that flames all over. It would be the last 939 anyone would need or want to play with.

You should.

It has a bit more of a learning curve than the Sapphire "Grouper", but IMHO it will clock better with a little TLC.

Seems with all things being equal the DFI will take your memory to a higher OC, granted it's a little spooky but with Tony's tweaks I have gone to a higher level.

[spajdr]

eva2000 : try to increase HT Bus Receiver to 31, i needed this to hit HT Link 1500Mhz

[eva2000]

thanks will try it later testing my Opteron 148 out now on the board.. seems both cpus of mine aren't clocking as well as on DFI NF4 series

DFI NF4 Ultra-D/UT SLI-DR max vs DFI CFX3200-DR UT max
Opteron 146 CAB2E 0540FPMW = 3150mhz vs 3100mhz
Opteron 148 CABYE 0543FPMW = 3100mhz vs 3060/3070mhz

[T2k]

What's the latest BIOS for CFX3200 w/ Corsair XMS4000PT modules (I have both 1GB and 2GB TWINX kits)?

[eva2000]

425 official bios is latest

Swapped to Opteron 148 CABYE 0543FPMW cpu as this will be the first cpu i test on Kayl's phase change unit on the CFX3200-DR platform. Might as well let them get acquainted with each other

Also decided to flash to the latest DFI official bios = 425 which loosened up the timings for Infineon based memory for better compatibility. This would be good for when i test 2GB OCZ PC4000EB Infineon based memory

Seems CFX3200-DR board needs a tad more vcore to do the same clock as on my DFI NF4 Ultra-D R.AD0 and DFI NF4 UT SLI-DR R.AB0 using same cooling and clocks 30-60mhz less benchable than the NF4 boards. It could be a memory setting i haven't yet tweaked since it's still early days.

But here's where i am at for 148 CABYE 0543FPMW (which did 3000Mhz at 1.536v on DFI NF4 series)

DFI UT CFX3200-DR R.A00 = 3000Mhz at 1.552-1.568v fluctuation


Cpuburn Session


Versus

DFI NF4 UT SLI-DR R.AB0 = 3014Mhz at 1.536v


Cpuburn Session


There's a 4-5C hotter cpu under load on DFI UT CFX3200-DR maybe due to the higher vcore applied ?

[spajdr]

still got shaky screen in 3D .. with new mobo and mastercard, i dont know what to do :-(

[eva2000]

Definitely liking this board now both my FX-57 and FX-60 clocked better just a tad although better H20 cooling might be the reason heh

DFI NF4
FX-57 = 3109-3120Mhz at 1.6-1.616v higher vcore didn't help
FX-60 = 3109Mhz at 1.6-1.616v higher vcore didn't help

DFI CFX3200-DR
FX-57 = 3137Mhz at 1.616-1.64v
FX-60 = so far at 3119Mhz at 1.568v

FX-60 @3119Mhz at 1.568v