E6750 & Gigabyte P35C-DS3R Overclocking Thread

[mrfatboy]

I have been talking to other people and it seems only the p35c-ds3r has the mem voltage problem. Everybody else's board actually does what it says




GIGABYTE TROUBLE TICKET UPDATE


More B.S. from those guys. I'm so disappointed in them. The person that is answering these "trouble tickets" is either an idiot or just does not care. Hey! It my even be a cover up for you consipiracy types

I wrote them back again and will update you later. BTW, my 1st question is about the memory voltage and the 2nd question is about the "cycle boot" problem.


software may not be properly indentifying the voltage, did you tried testing with an different spec of memory to see if it'll read the same.

Regarding the 2nd question, this was not an known issue.
If you wish to provide us with your system configuration and steps by steps procedure to duplicate your issue we will attempt to test it.

Thank you

 

[honolululu]

Originally posted by: mrfatboy


software may not be properly identifying the voltage, did you tried testing with an different spec of memory to see if it'll read the same.

:roll: Wonder what software was running on that multimeter?

 

[mrfatboy]

FYI, I had to reboot because I was installing some new software. When I did, my computer would not post again. I had to power it off at the PS and wait a bit. After 10 minutes, I power up at 2.66ghz. I then rebooted and when and reset to 3.6. Then it worked fine. I definately wasn't messing with the printer at the time This board has gremlins in it :evil:

Scary! tick tick tick. It's just a matter of time :Q

 

[paulmichaelkemp]

After running Orthos for 18hrs, I did a a super pi test and my results passed 15m 32.234s (32) so i assume everthing is ok . i rebooted and went back into bios, and in system health my DDR18v says " FAIL ".

It did this before. even when i changed my settings to default. so i updated my bios (F7). and it fixed the problem.

But what should i do now it's saying it again? do you think it's a false reading if i've passed these tests. or do i need to do something else?


I'm guess it's just saying failed because i have +.4 added, that is needed by these 2.2V DIMMS

 

[mrfatboy]

Congratulations you have successfully overclocked to 3.6

You can leave your current settings or try to lower either your Vcore to 1.45v or (g)mch back to normal. I'm not sure that you need both.

First test

Vcore 1.45625
(g)mch normal
Orthos stability test

2nd test

Vcore 1.45
(g)mch +.1
Orthos stability test


Tell us which test passes. Good luck.

Don't worry about the "fail" in PC health. It will always say that.

 

[paulmichaelkemp]

ah ok cheers mate! i tried the voltage at 1.4500 the firstime and it fail after 2 hours. i might just keep it as it is. i dont want to put to much stress on it. by lowering the volts.

my temps have been fine, just under 60 on load. in games.

idle cpu temps are 35c so i'm quite happy with that.


cheers


P.S - is there any need for me to run memtest?

 

[mrfatboy]

When your rig failed orthos at 1.45v did you have (G)mgh at +.1v? If not try that combo and test. But it sounds like your are done.

congrats.

 

[achanz]

Why Are you ppl stopping at 3.6? I'm also working on overclocking the 6750 but I found this forum discussion.. Let me know what you think.
http://forums.techpowerup.com/showthread.php?p=475188

 

[mrfatboy]

Originally posted by: achanz
Why Are you ppl stopping at 3.6? I'm also working on overclocking the 6750 but I found this forum discussion.. Let me know what you think.
http://forums.techpowerup.com/showthread.php?p=475188

I have the feeling that most people in this thread are looking for a high SAFE 24/7 Overclock on air. The 6750 seems to do 3.2 on stock volts then it need a little bump (1.45v) for 3.6. I know it can do more but at what cost. I'm not trying to break any records. I just want my computer to run 24/7 and last about 3-5 years or at least until I can't run a really cool game or app then I can upgrade

 

[achanz]

Has anyone here tried anything higher than 3.6 and it is no longer stable? Is 3.6 the max it goes with 100% stability? Has anyone played with higher volts and 3.7 for example... ?

 

[mrfatboy]

I don't think anybody on this thread had tried 3.7. I think 3.6 is the "best bang for the buck". Pushing your system beyond 3.6 just to get 1-3% extra performance is not really worth it in my opinion. You will never notice it.

My rig can already calcuate 1M PI in 14sec. I can care less if it does it in 13.9 All I need is to be faster then Sniperdawg and 5t3v0 :laugh:

 

[nCred]

I had it running 3.8 ghz at 1.475v stable, but i thought the northbridge got really hot and I had to run the fan on my AC7 Pro on max.
BTW I kept FSB and (G)MCH voltage at normal.

 

[SniperDaws]

Originally posted by: mrfatboy
I don't think anybody on this thread had tried 3.7. I think 3.6 is the "best bang for the buck". Pushing your system beyond 3.6 just to get 1-3% extra performance is not really worth it in my opinion. You will never notice it.

My rig can already calcuate 1M PI in 14sec. I can care less if it does it in 13.9 All I need is to be faster then Sniperdawg and 5t3v0 :laugh:

lmao, im quad core my time will come old girl

 

[honolululu]

Sometimes just for fun I'll crank up the volts and see how fast I can Pi. I usually don't go over 1.5V but last night it was freezing here at 73F/23C so I pushed it to up to 1.525V, +0.2V FSB and MCH, +0.3V DDR2. Don't think it did much, besides keep my feet warm. I can get to 480x8=3.840 at or below 1.5V, but couldn't get 485 to load Windows. I can do 500x7 without much trouble.

These are all with open case and a house fan blowing on the bridges, except the first pic here of the sigrig.
SuperPi=15.110s

Here are those 500x7 pics.
SuperPi =14.578s
Sandra Memory Bandwidth Oops. That's not the shot I wanted. Current setup in that pic would be 7*500 5-4-4-12. It was just shy of 8000 MB/s. It's actually a little colder this morning.

Okay. Here's 3.840 with auto RAM timings.
SuperPi=13.547s

3.840 with 5-5-5-15 SuperPi=13.422s
same in progress Pi in progress

3.840 with 5-4-4-12 SuperPi=13.391s

That Orthos start button should read restart. Total no go on Small FFT or Blend. Gromacs would run for 30 or so seconds.

Then this morning I read nCred's post about leaving FSB and MCH at normal. So I figured I'd give 485 another try this morning. 1.525V, Normal V FSB and MCH, +0.3V DDR2
SuperPi=13.375s That's a best for me.

EDIT: Just noticed I didn't try with 5-4-4-12 timings.
SuperPi=13.235s
All right! I had Gromacs going for 22s and was just going to take a screen shot and I got the red bar stop. Temps were up to 57C. I restarted Gromacs and 8 seconds later screen go black. Now I really gotta quit. Ambient is up to 79F/26C and rising.

 

[mrfatboy]

"Sometimes just for fun I'll crank up the volts and see how fast I can Pi. I usually don't go over 1.5V but last night it was freezing here at 73F/23C so I pushed it to up to 1.525V, +0.2V FSB and MCH, +0.3V DDR2. "


lol, 73F is freezing? Nice benchies tho. I wish I had an extra computer to play around with to do that higher end stuff.

 

[mrfatboy]

GIGABITE Trouble Ticket Update


Without knowing how to duplicate the issue how are we going to duplicate the issue?
Can you please assist us to replicating your issue?
We can test the memory to see what the board registers it at."


I told them that we cannot replicate the "cycle boot" problem but please read all of the forums to see what people are experiencing.

I also told them to please use Everest and Speedfan to measure memory voltages as well as reading directly from the motherboard.


I asked them both these questions last week

I will keep pushing them but they seem to be in denial mode

 

[ArchAngel777]

OP, your temperatures were almost extactly what mine were with my friends Q6750 @ 3.6Ghz (450x8 1.51v). I stressed it under dual prime and came up with 55 and 56 for each core under load after 30 minutes. It hit that temp after 5 minutes and held their steady.

Congrats, on the O/C BTW!

 

[Iksie]

Hello,
I'm following this tread every day and so i went oc-ing meself.
The strange thing with the gigabyte p35c ds3r is that when i do an overclock and i reboot the machine it wont boot at once but after a couple of times i will boot,but with the stock speeds and not the overclocked speeds.
Dont know what the deal about that,maby give more voltage?

i'm now running at 3.2 with this type of biossettings:

Bios Setup
-----------------
Robust Graphic Booster Auto
CPU Clock Ratio 8X
CPU Host Clock Control Enabled
CPU Host Frequencey (mhz) 400
PCI Express Frequency (mhz) 100
C.I.A.2 Disabled
!! Performance Enhanced with normal
System Memory Multiplier (SPD) [2.00]
Memory Frequency (Mhz) 800 900
High Speed Dram DLL Settings [option 1]
DRAM Timing Selectable (SPD) [manual]
CAS Latency Time 5 [4]
DRAM RAS# to CAS# 5 [4]
DRAM RAS# Precharge 5 [4]
Precharge delay(tRAS) 18 [12]
ACT to ACT Delay (tRRD) 3 [Auto]
Rank Write to READ Delay 3 [Auto]
Write to Precharge Delay 6 [Auto]
Refresh to ACT Delay 36 [0]
Read to Precharge Delay 3 [Auto]
tRD 7 [Auto]
tRD Phase Adjustment 0 [Auto]

System Voltage Control [auto]
Ddr2/ddr3 OverVoltage Control normal
PCI-E OverVoltage Control [Normal]
FSB OverVoltage Control [Normal]
(G)MCH OverVoltage Control [normal]
CPU Voltage Control [normal]
Normal CPU Vcore 1.35000V


I know that i can go more then 3.2 but when i increase the cpu voltage,then the system barely boots up and restarting a couple off times.
I know that de powersuply i got(sweex) is not an quality powersuply,but it can't be just that right?

Anybody got a clue how to run my machine faster?

 

[mrfatboy]

iksie,

Your computer is rebooting twice because it doesn't like the OC settings. It's a Gigabyte feature. I don't know your exact model number of your memory but I found something on Newegg that might be it. The memory voltage default is 2.0v. If you believe this board overclocks memory (i do for now) it defaults at 2.0v. However, adding +.1v to the memory also gives 2.0v.


Try increasing the ddr2 overvoltage to +.1v and see if that works.

System Voltage Control [auto]
Ddr2/ddr3 OverVoltage Control +.1v
PCI-E OverVoltage Control [Normal]
FSB OverVoltage Control [Normal]
(G)MCH OverVoltage Control [normal]
CPU Voltage Control [normal]
Normal CPU Vcore 1.35000V

 

[mrfatboy]

FYI, I found this thread in another forum. It seems the "cycle boot" problem is not just limited to the P35C.

Text

 

[jcossin]

Originally posted by: mrfatboy
FYI, I had to reboot because I was installing some new software. When I did, my computer would not post again. I had to power it off at the PS and wait a bit. After 10 minutes, I power up at 2.66ghz. I then rebooted and when and reset to 3.6. Then it worked fine. I definately wasn't messing with the printer at the time This board has gremlins in it :evil:

Scary! tick tick tick. It's just a matter of time :Q

Mrfatboy,

You definitely should upgrade your BIOS. F4g is old:

http://www.gigabyte.us/Support...el.aspx?ProductID=2524

When I upgraded my BIOS to the one that came out 9/7, those problems completely stopped for me. My board has been completely stable on boot now.

 

[mrfatboy]

@jcossin,

My bios is not that old plus gigabyte said they only made minor changes to it (ps2 keyboard). Remember, you and I have different boards. Since there seems to be some flakiness with the P35C, i'm taking a wait and see approach. There are still some people here running the F2 bios

 

[jcossin]

Originally posted by: mrfatboy
@jcossin,

My bios is not that old plus gigabyte said they only made minor changes to it (ps2 keyboard). Remember, you and I have different boards. Since there seems to be some flakiness with the P35C, i'm taking a wait and see approach. There are still some people here running the F2 bios

True, but since that BIOS came out on the same day as mine, I'd guess they have a common set of BIOS code that they tweak for each board and version separately.

Might be worth a try since you are not on the final F4 anyway and the my BIOS came out on the same day, had the same release notes, and did fix my problem. Easy enough to back down if you hit issues...

My guess is that they are not publishing all of the fixes they do when they do the release notes. I know in my software career they have been plenty of times I've fixed other things besides the main problem I was working on and didn't mention it...

 

[mrfatboy]

Gigabyte Trouble Ticket Update

It's taken 5 queston/answer sessions to get this response. They finally asked what Lot # I have :roll: Hmmmmm. Was there bad "lots" ? Only time will tell.

can you provide us with the LOT# you have
It's printed on the LPT port on the board.

 

[imported_Speeder]

I am having bad luck OCing.

I got a P35-DS3R (non C) F6 bios
E6750 G0 Stepping
2gb (2x1gb) Crucial Ballistix 4-4-4-12 ddr2-800 ram
ultima-90 HS (bundled thermal paste)
Scyche S-Flex 120mm F fan on CPU and E case fans.

Im running Vista 64-bit

I can do 8x400 4-4-4-12 but orthos freezes (Super PI runs). Games seem to run OK. Memtest passed with mem upto 475 with 4-4-4-12 (at 6x multiplier) with automatic Voltage. 485 freezes at that timing.

I can't do 490+ with any timing (at any DDR voltage setting... I get reboot cycles)
Orthos also freezes with 8x450 with voltage upto 1.48. I can't do 1.5 (reboot cycles) (I got to try some games at that setting... but it reboot cycles too on occasion and resets everything to default)

Maybe orthos just doesn't like my system period?

I have tried C1E, T2 (and whatever else) enabled & disabled. I have HPET 64-bit enabled but I don't think it does anything related to OC'ing.

I've tried DRAM option 1& 2. All perrmance stuff is standard or auto.

I have tried adding FSB and (G)MCH +.1V, didn't help

My goal is/was 450x8 with 4-4-4-12 timing, or maybe 515x7 with 5-5-5-15 timing.

The temp doesn't even seem to go that high, the system just freezes (orthos) or resets after reboot. Cores/CPU reach maybe... 40C at most.

Any suggestions?