[First Look] DFI 680i LT NF-680i LT SLI (56K Warning~~~)

[C-N]

Grinch check your numbers buddy your BIOS settings are 400 quad pumped to 1600fsb on the 8 multiplier but coretemp is showing 450 x 8 as its still reading your 6600's x9 multiplier & coming up with 450fsb to make 3.6GHz.

Having said that 3.2GHz is still a good clock.

Regards

Craig

[Grinch]

Grinch check your numbers buddy your BIOS settings are 400 quad pumped to 1600fsb on the 8 multiplier but coretemp is showing 450 x 8 as its still reading your 6600's x9 multiplier & coming up with 450fsb to make 3.6GHz.

Having said that 3.2GHz is still a good clock.

Regards

Craig

I noticed that as well...dunno how to fix it...but it is at 400x8 not 450x9

[C-N]

I noticed that as well...dunno how to fix it...but it is at 400x8 not 450x9

I don't think you can fix it because of how the programs (Orthos as well) read the multiplier use CPU-Z it will display the correct speed for you.

Craig.

Edit: Sorry I only commented because your sig showed 450 x 8 = 3.6GHz

[Grinch]

This is correct...*-)

[virtualrain]

We should have a database thread....

I'm willing to compile all the data into a spread-sheet and post it in the Tips/Tricks thread that's in my sig. (I'll rename it so it draws others as well).

...As long as everyone reposts their overclock on this page, so I don't have to comb back through 50 pages of this thread.

What I guess we should agree on is what settings to capture. Here is a first pass, including example data... am I missing anything critical? Keep in mind it can't be too much data or it won't be manageable. For BIOS voltages, let's use what is SET in BIOS to keep things consistent.

Processor: X3220
RAM: Mushkin PC2-8500 2x2GB
BIOS ver: 521
OC: 400x9
Stability: PRIME 8+hrs
Cooling: Water (Apogee GTX)
Vcore: 1.6V + 0% (strong)
Vdimm: 2.34V (weak)
NB: 1.65V
SB: 1.62V
LDT: 1.52V
VTT: 1.6V
FSB: Linked/Sync
RAM Clk: 400
Timings: 4-3-3-8 1T


Let me know your thoughts.

[Grinch]

I'm willing to compile all the data into a spread-sheet and post it in the Tips/Tricks thread that's in my sig. (I'll rename it so it draws others as well).

...As long as everyone reposts their overclock on this page, so I don't have to comb back through 50 pages of this thread.

What I guess we should agree on is what settings to capture. Here is a first pass, including example data... am I missing anything critical? Keep in mind it can't be too much data or it won't be manageable. For BIOS voltages, let's use what is SET in BIOS to keep things consistent.

Processor: X3220
RAM: Mushkin PC2-8500 2x2GB
BIOS ver: 521
OC: 400x9
Stability: PRIME 8+hrs
Cooling: Water (Apogee GTX)
Vcore: 1.6V + 0% (strong)
Vdimm: 2.34V (weak)
NB: 1.65V
SB: 1.62V
LDT: 1.52V
VTT: 1.6V
FSB: Linked/Sync
RAM Clk: 400
Timings: 4-3-3-8 1T


Let me know your thoughts.

Looks good to me...

[Pete@X]

I'm willing to compile all the data into a spread-sheet and post it in the Tips/Tricks thread that's in my sig. (I'll rename it so it draws others as well).

...As long as everyone reposts their overclock on this page, so I don't have to comb back through 50 pages of this thread.

What I guess we should agree on is what settings to capture. Here is a first pass, including example data... am I missing anything critical? Keep in mind it can't be too much data or it won't be manageable. For BIOS voltages, let's use what is SET in BIOS to keep things consistent.

Processor: X3220
RAM: Mushkin PC2-8500 2x2GB
BIOS ver: 521
OC: 400x9
Stability: PRIME 8+hrs
Cooling: Water (Apogee GTX)
Vcore: 1.6V + 0% (strong)
Vdimm: 2.34V (weak)
NB: 1.65V
SB: 1.62V
LDT: 1.52V
VTT: 1.6V
FSB: Linked/Sync
RAM Clk: 400
Timings: 4-3-3-8 1T


Let me know your thoughts.

I would add as extra point (if not mentioned extra, ppl maybe don`t get it )

CPU Drive Strenght: weak/strong <-- specially usefull for Quads
DRam Drive Strenght: weak/strong <-- weak keeps the rams alive

and the GTLs

Time to get the boards finaly mounted - testing is over, h20 parts are here ... and the new Mushkin 996561 need a new home

[virtualrain]

I would add as extra point (if not mentioned extra, ppl maybe don`t get it )

CPU Drive Strenght: weak/strong <-- specially usefull for Quads
DRam Drive Strenght: weak/strong <-- weak keeps the rams alive

and the GTLs

Time to get the boards finaly mounted - testing is over, h20 parts are here ... and the new Mushkin 996561 need a new home

I already had drive strengths associated with the relevant voltage as highlighted below.

I was torn on GTL's... they are totally unique from one combo to the next and one's man pill can be another man's poison with GTL. At any rate, there's no harm in sharing. So let's do this...

Processor: X3220
RAM: Mushkin PC2-8500 2x2GB
BIOS ver: 521 2C
OC: 400x9
Stability: PRIME 8+hrs
Cooling: Water (Apogee GTX)
Vcore: 1.6V + 0&#37; (strong)
Vdimm: 2.38V (weak)
NB: 1.65V
SB: 1.62V
LDT: 1.52V
VTT: 1.6V
GTL: 195/195/95
FSB: Linked/Sync
RAM Clk: 400
Timings: 4-3-3-8 1T

If you want to start putting your settings in this format, go ahead... if we change the settings based on further feedback, you can always come back and edit your post.

[Grinch]

Ok here is mine:

Processor: E6600 Conroe
RAM: Corsair XMS 6400C4 DDR2 800
BIOS ver: 521 2C
OC: 400x8
Stability: PRIME 4+hrs
Cooling: Water (D-Tek Fuzion)
Vcore: 1.28125V + 0&#37; (weak)
Vdimm: 2.01V (weak)
NB: 1.44V (Water Swiftech MCW30)
SB: 1.56V
LDT: 1.22V
VTT: 1.21V
GTL: Disabled
FSB: Linked/Sync
RAM Clk: 400
Timings: 4-4-4-12 2T

[Jobeo]

ya i will post my settings in a bit once i get them stable again.. i had a 400x8 on a qx6700 which proved OCCT 30 min stable a couple times... then a couple days later i'm playing supcom and it crashes. i think hrmm this is strange

so i run OCCT and it fails. i tried tweaking a couple voltage settings and it still fails. its strange cause i hadn't changed a damn thing from the time it was passing until it started failing. no bios updates or anything.

i guess i've learned that OCCT isn't a reliable stability test. although it seems strange that it passed several times then a couple days go by and its all fail fail fail.

[Grinch]

i still like to use prime95 using small fft's

[virtualrain]

This is the easiest program to use for stress-testing quads... http://www.mersenne.org/gimps/P95V252.ZIP

Use small FFT's (which fit in cache) to stress your CPU OC and use Blend to test NB/RAM.

[Borgod]

I'll post mine tonight. Im stuck on 3550Mhz with my QX6700 and having a real hard time getting it Prime Stable 8hrs on 3600Mhz. 1.5hrs is the longest Ive gone, although this seams to be stable gaming for many hours and video encoding. I still have to increase GTL as far as virtualrain has gone though, so a bit of headroom left.

Am I mad for considering a 3rd loop to cool NB and PWM's? lol

[Grinch]

not if you are looking for maximum performance...

[GNU]

Very interesting posts and I have some questions:

1.) What's the sense of SPP and PCIe oc? I tried PCIe 130 and SPP 250 and I never seen any changes (faster or stable etc.).

2.) The vdroop is extremely and what's the bios settings for a normal voltage (ex: 1,7 in bios = real 1,5v) I haven't a meter and hope you can give me some informations

I'm on 3200 (for 24/7) @950 1T stable

[owcraftsman]

Hey Grinch,
You bet It wouldn't be pretty if I caught the thief and yes I plan to water cool.
I'm still trying to make up my mind about case and cooling gear.

Parts I'm considering:
SWIFTECH APOGEE GTX EXTREME PERFORMANCE CPU WATER BLOCK (1/2" BARBS)
HARDWARE LABS BLACK ICE PRO II 240 RADIATOR (1/2" BARBS)
DANGER DEN GPU-6200 GPU WATER BLOCK FOR NVIDIA 8800GTS (1/2" BARBS)
SWIFTECH MCW30 CHIPSET COOLER FOR AMD AND INTEL (1/2" BARBS)
AquaXtreme™ 50Z-DC12 12V DC Pump
TYGON R-3603 7/16" ID X 5/8" OD TUBING
Fluid XP+ Extreme - UV Midnight Blue
COOLER MASTER STC-T01-UW1 SILVER CM STACKER with STF-B01-E1 CROSS FLOW FAN ???? I wish I could find an aluminum as good


Okay, I'm sure everybody has opinions about these choices and I WELCOME COMMENTS but please, if you want to, PM me instead of taking this thread off course. TY

Virtualrain, When/if I ever get all my parts together I'd be happy to repost here results. I was thinking more along the lines of a data base with a standardized format. Let's say minimum 8 hrs prime (blend & a 10 priority) a 3DMark06 run and 3 instances of CPUZ (mainboard memory and memory spd)
Everyone on the same page. I'm not saying anyone would fake results but I've seen and read some outrageous comments at times. Non the less I'd be on board with your plan. I've read and book marked your site and Ty very much for the hard work. Very professional and well put together.

[virtualrain]

Very interesting posts and I have some questions:

1.) What's the sense of SPP and PCIe oc? I tried PCIe 130 and SPP 250 and I never seen any changes (faster or stable etc.).

2.) The vdroop is extremely and what's the bios settings for a normal voltage (ex: 1,7 in bios = real 1,5v) I haven't a meter and hope you can give me some informations

I'm on 3200 (for 24/7) @950 1T stable

I can't help you with #1, but my guess is that overclocking these will only help if you can somehow saturate the HT bus or the PCIe bus. Neither is likely so that's why you don't see any improvement.

With #2, I did some measurements... see my findings near the bottom of this article (under BIOS Voltage Settings).

[virtualrain]

Virtualrain, When/if I ever get all my parts together I'd be happy to repost here results. I was thinking more along the lines of a data base with a standardized format. Let's say minimum 8 hrs prime (blend & a 10 priority) a 3DMark06 run and 3 instances of CPUZ (mainboard memory and memory spd)
Everyone on the same page. I'm not saying anyone would fake results but I've seen and read some outrageous comments at times. Non the less I'd be on board with your plan. I've read and book marked your site and Ty very much for the hard work. Very professional and well put together.

I agree this would be nice, and I considered this, but the problem is that you can't force people to adhere to any kind of stability requirements... unfortunately. Given the community for this board is small at best, I thought the database would be more useful with more entries and everyone could just add their stability test along with their settings and then you can judge whether you want to ignore their entry or not based on that.

[GNU]

@vr: Thank you

So I have made some benches.

What do you think?

Watertemp is 33- 34°C @idle (under load 35-38°C)
CPU Temp is 55°C @idle (under load ~60°C)

I don't know, it's an acceptable oc for 24/7 or too hard (too hot and too much voltages)?

[virtualrain]

That looks great... awesome memory bandwidth!

I find it hard to believe that your CPU temps only fluctuate by 5-deg between idle/load. Mine go from about 45 at idle to 65 at load on a quad with 1.55V... with a dual-core you should see much lower idle temps than that.

Anyway, your voltages are fine (except maybe your VDimm which I can't read... is that blotched out on purpose?)

[GNU]

Thank you,

Yes, it's hard to believe but it's maybe a sensor problem.
Maybe it's the sun (My PC is near the window and the day is hot (35&#176;C in the sun). The water was a little bit heated by the sun :/

Vdimm = Bios 2,22 set /get 2,18 something. Sorry about the pic. I will reupload a new pic. It's not blotched out, it's a mistake with paint :p

Do you mean the system is okay with this temp for 24/7 with max 60&#176;C under load?

OCCT gives me an error (temp error = CPU is overheated) and I don't know, shall I drop the FSB or whatever to get lower temps?

I will wait until a earlier night to retest the system with OCCT (only a 30 min test to see, what happens).

Edit: I had restart the system (start temp was 45&#176;C after 15min now 50&#176;C @idle). Pic was uploaded. Watertemp 31&#176;C and no Sun is cooking the water but a hot air (30&#176;C)

Edit2: After a 3DMark05 Bench with 11425 Points SG shows CPU 52&#176;C NB 48 and Sys 40.

Edit3: After a long idle modus: CPU 56&#176;C , NB 50, Sys 41 and water = 34

[MikeMK]

For those of you in the UK, OCUK have the DFI's in stock at last! Grab one quick though as they seem to be going fast:

http://www.overclockers.co.uk/showpr...atid=5&subcat=

After much deliberation, but weds i will have both an EVGA AND a DFI sat here on my desk LOL! gonna try both and drop what doesnt work for me. The biggest prob I see is with my Galaxy... fingers crossed, the DFI will be ok

[virtualrain]

Thank you,

Yes, it's hard to believe but it's maybe a sensor problem.
Maybe it's the sun (My PC is near the window and the day is hot (35°C in the sun). The water was a little bit heated by the sun :/

Vdimm = Bios 2,22 set /get 2,18 something. Sorry about the pic. I will reupload a new pic. It's not blotched out, it's a mistake with paint :p

Do you mean the system is okay with this temp for 24/7 with max 60°C under load?

OCCT gives me an error (temp error = CPU is overheated) and I don't know, shall I drop the FSB or whatever to get lower temps?

I will wait until a earlier night to retest the system with OCCT (only a 30 min test to see, what happens).

Edit: I had restart the system (start temp was 45°C after 15min now 50°C @idle). Pic was uploaded. Watertemp 31°C and no Sun is cooking the water but a hot air (30°C)

Edit2: After a 3DMark05 Bench with 11425 Points SG shows CPU 52°C NB 48 and Sys 40.

Edit3: After a long idle modus: CPU 56°C , NB 50, Sys 41 and water = 34

60-deg under load is fine.

If your system is not stable, you probably need more voltage... you still have a bit of room to move there... try increasing your CPU and NB voltage a notch or two and test it... stress test it at night when it's cooler.

[virtualrain]

For those of you in the UK, OCUK have the DFI's in stock at last! Grab one quick though as they seem to be going fast:

http://www.overclockers.co.uk/showpr...atid=5&subcat=

After much deliberation, but weds i will have both an EVGA AND a DFI sat here on my desk LOL! gonna try both and drop what doesnt work for me. The biggest prob I see is with my Galaxy... fingers crossed, the DFI will be ok

I'll be keen to hear your results with both boards. I was tempted to do that same evaluation but opted to just grin and bear it with the DFI!

[MikeMK]

Yup, well i think it will take me longer to get to grips with the DFI and the EVGA, still it will be interesting to see how it all pans out and if the pain is worth the gain!