Gigabyte P965-P35 X38 overclocking and BIOS tweaking Guide

[amd7674]

Thanks BIll for this excellent guide !!!! ;-) I'll post my o/c results soon.

 

[amd7674]

I did some initial o/c last night and the best 'stable' I could do was 3Ghz 9 * 334 , I've tried ( 9 * 356) but I was rebooting while stressing the system, no matter how much voltage I'v e applied up to 1.45V....LOL My memory was running in 4-5-4-11 @ 2.10V, I've tried 5-5-5-15 @ 2.10 without luck...

I'll try 8 * 400 tonight.... How should I setup my memory (timings/voltage) for it?

Any other sugestions?

TIA

From top of my head the settings were:

CPU Clock Ratio (Note)____________ [X9] <<<----CPU Multiplier
CPU Host Clock Control_ [Enabled]
CPU Host Frequency (MHz)__________ [334] <<<----FSB Speed (Front Side Buss)
PCI Express Frequency (Mhz)_______ [102]
C.I.A. 2__________________________ [Disabled]
System Memory Multiplier (SPD)____ [2.00]
Memory Frequency (Mhz) 668
DRAM Timing Selectable_______ SPD __ [manual]
CAS Latency Time_____________ [5] [4]
Dram RAS# to CAS# Delay______ [5 [5]
Dram RAS# Precharge Delay_____[5] [4]
Precharge Delay (tRAS)________[15] [11]
ACT to ACT Delay (tRRD)_______4 _____[auto]
Rank Write to READ Delay______3 _____[auto]
Write to Precharge Delay______6 _____[auto]
Refresh to ACT Delay________42 ______[0]<-
Read to Precharge Delay_______4 _____[auto]
Memory Performance Enhance__________ [Normal]
High Speed DRAM DLL Settings________ [Option 1]
System Voltage Control____ [Manual]
DDR2 OverVoltage Control__ [+0.3V]
PCI-E OverVoltage Control_ [+0.1V]
FSB OverVoltage Control___ [+0.1V]
(G)MCH OverVoltage Control [+0.1V]
CPU Voltage Control_______ [1.38125V]

 

[Foz2001]

I have had my e6400 OC'd to 2.8ghz for awhile now. Setup in mobo was:

FSB: 400
Ram: 2 (1:1)
Multiplier: 7
Voltage: 1.325


Now, I just got my 120 extreme, which means my temperatures are a lot lower so I decided to push this chip more. So my initial thought was to just increase the multiplier to 8. No luck. I then pushed the voltage up to 1.4 and still no luck. What is up with that? 1.4v should be able to make it run at 3.2GHZ no?. Any thoughts?

Foz
____________________________________

 

[GwarGor]

My e6400 hit a wall @ just over 3.1ghz, so I backed it down to 3Ghz just to be safe and 3Ghz is still fast enough IMO. Try 8x376 and that will get you to 3Ghz and then you could try to increase it step by step from there. I also didn't have to up the voltage anymore than this guide suggested to do in the first place to achieve it.

 

[jtsarnak]

I am suddenly having issues with this board and rebooting. It appears to be shutting down but then the monitor goes black and the system lights stay on. I then turn off the power supply switch and when I turn it back on, the computer doesn't turn on when I hit the power switch. If I walk away from it for 10-15 minutes, I will find it turned on already.

Here are my specs:
Vista Home Premium 32-bit
E6420
DS3 rev 3.3 Bios F12
XFX Geforce 8800GTS 320MB
Patriot Extreme 800 1GBx2

I had the system overclocked at 8x400 for awhile, but then when I had issues I removed the overclock and still get the same problem.

I read about warm boot issues and tried my PCI-e at 101 and 105 Mhz, added +.1 volts to everything for stability, but still have the issue.

I also recently switched the HDD from the purple port to the yellow and disabled the AHCI options in the BIOS. Could this have something to do with it?

 

[Jaybe]

I'm new to OC but wanted to give it a shot based on the hardware I have purchased over the last 6 months. I thought I could get a little extra horsepower out of my chip - no gaming, just office type applications. Eventually, I might even get a game or two

My setup is:
Vista Ultimate 32 bit
965P-DS3 Rev 1.3 with F11 bios
E6600
XFX Geforce 7600 GS with latest drivers (169.25)
Corsair xms2 (2 x 1GB)
Zalman 9500 cooling

I started by stepping though the tutorial to get my baseline, however when I do this, the system seems to run very slow. It is slow to load at startup, slow to load programs, and seems to be chugging along (at a slow pace). Prior to this, I would startup and the welcome screen would pop up within seconds, now it takes about 45 seconds. When opening CPU-z, Coretemp, or any program, it takes about 5 seconds for the UAC to popup asking if it is okay to run this program.

Also, I get "Display driver nvlddmkm has stopped responding and has been recovered." I'm thinking that this could be the root of the problem, but I'm not sure.

When I load the fail safe values in the bios, everything runs quickly again, although the processor is underclocked at 1.6. This thing should be very fast and I can't seem to get even the baseline to work. Any ideas on where to start troubleshooting? Could it be something with the video drivers? Why does the fail safe run so fast compared to the optomized defaults?

Thanks in advance for any suggestions,
John

 

[rbreding]

I have a perplexing problem and have worked on it for a couple weeks now and need assistance.

I had been running on a E6600 at 410x8 for the longest time (months). Did a normal reboot one day and the system did the double start (as if something were wrong) and reset the overclock. No biggie I thought, just go into the Bios and load back up my overclock profile. No dice. No matter what setting/voltage I have tried since then it won't get past the memory test before it does the double boot and resets the settings.

I have even tried just to manually enter in the stock speeds instead of choosing auto to see if that worked (like it would before) and nothing. The only thing that will completely boot is if the FSB is set to the default settings.

I even tried another CPU (e6550) to see if I cooked something in the processor....no dice either (meaning no overclock in the bios). The weird thing is that I can use Easytune (last resort) to overclock and it works fine, however the settings will not survive a reboot. But it does overclock fine.

I need to retain what hair I have left, so help me


EDIT: I even tried different ram....same result (ram test fine at all speeds I have tried with memtest)

Just as a followup to my problem above. I tried another power supply and the problem persisted. I also tried an e4300. So that is three processors that are exhibiting the same problem.

Possible board problem ? Could it be RMA'd, and under what auspice ?

 

[BillParrish]

I did some initial o/c last night and the best 'stable' I could do was 3Ghz 9 * 334 , I've tried ( 9 * 356) but I was rebooting while stressing the system, no matter how much voltage I'v e applied up to 1.45V....LOL My memory was running in 4-5-4-11 @ 2.10V, I've tried 5-5-5-15 @ 2.10 without luck...

I'll try 8 * 400 tonight.... How should I setup my memory (timings/voltage) for it?

Any other sugestions?

TIA

From top of my head the settings were:

CPU Clock Ratio (Note)____________ [X9] <<<----CPU Multiplier
CPU Host Clock Control_ [Enabled]
CPU Host Frequency (MHz)__________ [334] <<<----FSB Speed (Front Side Buss)
PCI Express Frequency (Mhz)_______ [102]
C.I.A. 2__________________________ [Disabled]
System Memory Multiplier (SPD)____ [2.00]
Memory Frequency (Mhz) 668
DRAM Timing Selectable_______ SPD __ [manual]
CAS Latency Time_____________ [5] [4]
Dram RAS# to CAS# Delay______ [5 [5]
Dram RAS# Precharge Delay_____[5] [4]
Precharge Delay (tRAS)________[15] [11]
ACT to ACT Delay (tRRD)_______4 _____[auto]
Rank Write to READ Delay______3 _____[auto]
Write to Precharge Delay______6 _____[auto]
Refresh to ACT Delay________42 ______[0]<-
Read to Precharge Delay_______4 _____[auto]
Memory Performance Enhance__________ [Normal]
High Speed DRAM DLL Settings________ [Option 1]
System Voltage Control____ [Manual]
DDR2 OverVoltage Control__ [+0.3V]
PCI-E OverVoltage Control_ [+0.1V]
FSB OverVoltage Control___ [+0.1V]
(G)MCH OverVoltage Control [+0.1V]
CPU Voltage Control_______ [1.38125V]

with a 2.0 memory mulitplier and as sub 400 FSB your memory should not be an issue. However take it completely out of the picture while you work on the cpu. Set the memory to 5 5 5 15 and voltage to 1.9V.

To address the low FSB raise the MCH voltage to +.2 and if there is not one, glue a fan to the NB. Cannot guarentte that will help but it cant hurt and is the first thing to try. The MCH/NB gets OCed just like the cpu so keeping it cool is a key to higher FSB speeds.

 

[BillParrish]

I have had my e6400 OC'd to 2.8ghz for awhile now. Setup in mobo was:

FSB: 400
Ram: 2 (1:1)
Multiplier: 7
Voltage: 1.325


Now, I just got my 120 extreme, which means my temperatures are a lot lower so I decided to push this chip more. So my initial thought was to just increase the multiplier to 8. No luck. I then pushed the voltage up to 1.4 and still no luck. What is up with that? 1.4v should be able to make it run at 3.2GHZ no?. Any thoughts?

Foz
____________________________________

It all depends on luck of the draw on cpus. I would try (and watch temps like a hawk and if it works the first thing I would do is determine the least amount of Vcore that will run it stable) 1.45V set mch +.2, fsb at 360 mulitplier of 8 make sure you memory multiplier stays at 2.0 and go up 10MHz a try and find out what is limiting you and how high you can get. Glue a 40mm fan onto the MCH/NB.

 

[BillParrish]

I'm new to OC but wanted to give it a shot based on the hardware I have purchased over the last 6 months. I thought I could get a little extra horsepower out of my chip - no gaming, just office type applications. Eventually, I might even get a game or two

My setup is:
Vista Ultimate 32 bit
965P-DS3 Rev 1.3 with F11 bios
E6600
XFX Geforce 7600 GS with latest drivers (169.25)
Corsair xms2 (2 x 1GB)
Zalman 9500 cooling

I started by stepping though the tutorial to get my baseline, however when I do this, the system seems to run very slow. It is slow to load at startup, slow to load programs, and seems to be chugging along (at a slow pace). Prior to this, I would startup and the welcome screen would pop up within seconds, now it takes about 45 seconds. When opening CPU-z, Coretemp, or any program, it takes about 5 seconds for the UAC to popup asking if it is okay to run this program.

Also, I get "Display driver nvlddmkm has stopped responding and has been recovered." I'm thinking that this could be the root of the problem, but I'm not sure.

When I load the fail safe values in the bios, everything runs quickly again, although the processor is underclocked at 1.6. This thing should be very fast and I can't seem to get even the baseline to work. Any ideas on where to start troubleshooting? Could it be something with the video drivers? Why does the fail safe run so fast compared to the optomized defaults?

Thanks in advance for any suggestions,
John

Frankly , I dont have a clue, adjusting speeds etc. in the bios should not affect the video driver like that. Troube is I do not use Visa and have no idea what kind of strange things are going on. I sounds like a video driver issue with a corrupted file. To eliminate that possbility I would remove the driver from add/remove programs, remove the video card from Device manager, run drivercleaner (free version), reboot, on reboot cancel any "new hardware found msgs" and then reinstall the driver from a fresh download. This is for XP not sure how to translate into vista. but it is a software problem. Perhaps a crash while playing with the OCing corrupted the file /shurg. But I have never heard of a driver file acting like that just by changing cpu speed etc. So a new clean driver install should fix you up.

 

[jtsarnak]

Any ideas on my issue Bill? Any help would be greatly appreciated.

 

[Torpin]

I have a perplexing problem and have worked on it for a couple weeks now and need assistance.

I'm interested in what you find about this, rbreding. I have similar issues with my P35-DS3L board. Although in my case I only had the board for a week (stable OC and all) before the first "double boot" occured and now it's very flaky. It seems almost ok if I leave the system voltages etc on auto. But any time I try and adjust settings manually to reduce heat, bigtime problems occur. Even doing nothing but enabling FSB freq control but leaving it on the default for my processor can make it reset!

 

[rbreding]

Only thing I haven't tried is flashing to bios to a previous rev....might try that next.

 

[Torpin]

Well I updated to a newer bios but it made no difference. The problem seems to be getting worse.

After googling over a week it seems both these chipsets can suffer the problem. And there is no real fix. Except to get another one and cross fingers. Just luck of the draw really.

 

[mbrockma]

I am trying to overclock my new computer (first desktop I've had) and I ran into some trouble. I followed the instruction on the front page, after setting the baseline values and loading windows, I opened Core Temp and CPU-z. Then I opened Prime95 25.2 (the Orthos from the link on the front page would not work for me, It gives an error when you try to open it, so instead I downloaded it from here...http://www.diy-street.com/forum/showthread.php?t=61194, fourth link down).

I select blend test then start, and it immediately gives an error message, something along the lines of "Fatal Error: rebounding was .5, expected to be less than .4" and hardware error, the test stops right away.

I tried running memtest to see if my ram is good, but for some reason it will not work. The CD is in the drive and the bios is set to boot from the CD first, it gets to a point where it should say press an key to boot from cd, but instead it just says "Loading............" then does nothing. I checked the CD in my laptop and it works. The optical drive worked fine to install vista and I even did memtest a couple weeks back using a Linux disc (this disc no longer works either, I can actually get it to load the linux menu and select test memory, then it just goes blank).

So, what do I do next?

I have a P35C-DS3R rev 1.1 mobo, e6750, OCZ 4GB DDR2-800 (5-4-4-15 2.1V), I'm new to this, so if I did something wrong, let me know so I don;t do it again.

 

[BillParrish]

OCZ 4GB DDR2-800 (5-4-4-15 2.1V),

This is only a guess, but I would immedately set your memory volate to 2.1V if you have not already done so. Set the MCH voltage to +.2 to help drive that 4GB of memory.

The second thing to try is to remove ram and just run 2 GB.

I will need detailed information on exactly what all the MIT menu items are set at to really make a good guess at what is going on.

I will fix the orthos link. Thanks.

 

[Redshirt #24]

Quick thread hijack, but it is BIOS-related...I bought a P965-DS3 (revision 1.0--at least it was relatively cheap, but I know...) a few weeks ago as a spare/replacement for my eVGA 680i. In a nutshell: I have no idea which BIOS is installed on it, but the previous owner said he hadn't updated it in over a year. So I wonder: since my only CPU is the Q6600 I'm using now, will I even be able to boot with the thing if it's running an older BIOS?

 

[Evilkiksass]

I followed your guide and some advice from friends when overclocking my Q6600 and was able to push it to 3.0Ghz stable by just upping the FSB and dropping the memory multiplier. As I attempted to push further I had to up the voltage and was able to achieve 3.2Ghz stable with the included bios settings. The odd thing is that if I push it to 3.3 Ghz and up the voltage as high as 1.525v to the CPU it still will not hold stable. The memory should not be having an issue with 816Mhz, and I can tell that the NB is cool enough by touching it. At the moment I am trying to figure out why I can not manually set the memory timings, and why this CPU does not want to go any further. The temps are in the low 80s (CoreTemp) under full load on all 4 cores, I have a tjunction of 100c. One thing I did notice was that in CPU-Z the voltage is MUCH lower then what I set. It is set to 1.475 in bios, it is 1.425 idle and it is 1.362 under load. I have read the article on Anandtech about how Vdroop works however I do not see why it would drop the power low enough that it would affect the performance. Any advice would be appreciated.

Hardware specs:
GA-965P-DS3 Rev 3.3 Bios Version F12
Intel Q6600
2 x Corsair CM2X-1024-6400
Antec Neo HE 650 Watt Powersupply
I have a custom made water cooling system on the cpu.
And a 40mm fan I pulled out of a 1u server case on the NB.

CPU Clock Ratio (Note)____________ [X8] <<<----CPU Multiplier
CPU Host Clock Control_ [Enabled]
CPU Host Frequency (MHz)__________ [400] <<<----FSB Speed (Front Side Buss)
PCI Express Frequency (Mhz)_______ [102]
C.I.A. 2__________________________ [Disabled]
System Memory Multiplier (SPD)____ [2.00]
Memory Frequency (Mhz) 800
High Speed DRAM DLL Settings________ [Option 1]
DRAM Timing Selectable_______ SPD __ [Auto] (I had to set this, even with all stock on manual it will not boot into windows)
CAS Latency Time_____________ [5] [5]
Dram RAS# to CAS# Delay______ [5 [5]
Dram RAS# Precharge Delay_____[5] [5]
Precharge Delay (tRAS)________[18] [18]
ACT to ACT Delay (tRRD)_______3 _____[auto]
Rank Write to READ Delay______3 _____[auto]
Write to Precharge Delay______6 _____[auto]
Refresh to ACT Delay________42 ______[42]
Read to Precharge Delay_______3 _____[auto]
Memory Performance Enhance__________ [Normal]
System Voltage Control____ [Manual]
DDR2 OverVoltage Control__ [+0.4V] (Corsair claims this memory can run at 4-4-4-12 @ 2.2v)
PCI-E OverVoltage Control_ [+0.1V]
FSB OverVoltage Control___ [+0.3V]
(G)MCH OverVoltage Control [+0.2V]
CPU Voltage Control_______ [1.47500V]

 

[mbrockma]

This is only a guess, but I would immedately set your memory volate to 2.1V if you have not already done so. Set the MCH voltage to +.2 to help drive that 4GB of memory.

The second thing to try is to remove ram and just run 2 GB.

I will need detailed information on exactly what all the MIT menu items are set at to really make a good guess at what is going on.

I will fix the orthos link. Thanks.

EDIT: I just set the MCH to .2V instead of 0.1V and it works. thanks for the help.
EDIT 2: I tried continuing on with the guide, and overclocking the fsb to 375 (instead of 333) and I got the same error immediately after starting prime95. I went back to the baseline (with .2 MCH) and tried prime95 again, now it does not work again, I get the error right away.

See MIT menu settings below. Prime95 will run with only 1 stick of ram (I tried both) but not all 4GB at the same time. I get this message as soon as the test starts:

"FATAL ERROR: Rebounding was 0.5, expected less than 0.4
Hardware failure detected, consult stress.text file.
Torture test ran 0 minutes - 1 error, 0 warnings.
Work thread stopped"

MIT:
Robust Graphics Booster - auto
CPU clock ratio - 8x
cpu host clock control - enabled
cpu host frequency - 333
pci express frequency - 102
cia2 - disabled
system memory multiplier - 2.00
memory frequency - 800 667
high speed dram dll settings - option 1
performance enhance - turbo
dram timing selectable - manual
cas latency time - 5
dram ras# to cas# delay - 4
dram ras# precharge - 4
precharge delay (tras) - 15
act to act delay -auto
rank read to read delay - auto
write to precharge delay - auto
refresh to act delay - 0
read to precharge delay - auto
trd - auto
trd phase adjustment - auto

system voltage control - manual
ddr2/ddr3 overvoltage control - +.3V
pci-e - +.1V
FSB - +.1V
MCH - +.2V
CPU voltage control - 1.375V
normal cpu vcore - 1.35V

 

[chrispycrunch]

This is such an informative thread - I printed your writeup and read through (skimmed) the rest of the posts. Any comments for E6x50 CPUs? i.e. the CPU / Memory Divider / FSB ratios?

Even though I have the DS3R (P35), this thread is very much applicable for me.

 

[silverspade14]

Hello.

Yesterday I used this guide to overclock my C2D e6400.I got it up to 2.6 GHz but when I read the thing about how the Northbridge is also overclocked and also needs heating, I stopped there. My computer was fine all last night. I woke up this morning and turned it on, surfed around a bit, and I leave for a few minutes. I come back and my computer had restarted. I figured it was an auto update. I surfed some more and started to watch a movie and all of a sudden the screen goes blue and then shuts down. I couldnt read what the blue screen said. Anyways I reset my bios to the defaults. I was wondering, is this a problem with my memory voltage? Was my Northbridge too hot? Any help you can give me would be nice.

(I just ordered a fan for my Northbridge yesterday, before the crashing started)

Edit: Btw my core temps werent getting very high, about 39-42 C idle and when I was running the test on Orthos it only got up to 48C

 

[GwarGor]

I also have an e6400(DS3 P965 r3.3) and I got it to 3Ghz and don't use a NB fan but my CPU heatsink/fan has fins that deflect air to blow on it. The thermal paste used on the NB isn't always applied evenly and that could be part of your problem too. If you have some good thermal paste you might want to reapply some on the NB. Also what board do you have?

 

[silverspade14]

I also have an e6400(DS3 P965 r3.3) and I got it to 3Ghz and don't use a NB fan but my CPU heatsink/fan has fins that deflect air to blow on it. The thermal paste used on the NB isn't always applied evenly and that could be part of your problem too. If you have some good thermal paste you might want to reapply some on the NB. Also what board do you have?

I have the same board as you. I don't know if its r3.3 though.

 

[mau108]

just an update, still running my e6400 with 1.0 rev mobo (ds3) at 3.2ghz with a big fat thermalright ultra 120 fan and antec nine hundred case .. ubber stable...heck I think more stable then my stock amd htpc

Been like this for 5-6 months now

 

[silverspade14]

So does anyone have any ideas about my problem? Is it my northbridge overheating?

 

[BillParrish]

So does anyone have any ideas about my problem? Is it my northbridge overheating?

The long and the short of it is, put the fan on the NB, $5 and couple of dabs of RTV silicone adhesive and you do not have to worry about it. There is no way to determine exactly what your issue may be as there is insufficient information. What memory do you use and what is the manuf suggested voltage ? What were the FSB and multiplier setting etc. etc. It seems the cpu Heatsink/fan is working fine so that is probally not the issue.

 

[bardos]

first let me thank bill parrish for the guide on the first page of this thread for his help in setting up my board and guiding me thru the myriad list of unknown bios terminology.

On my testing bench at the moment i've got a ga-p35-d3sl rev 2 board, running an e4300 CPU at 335fsb (3.0Ghz) at stock voltage (1.325.) and a rosewill RCX-Z775-EX HSF.

2gb of mushkin pc6400 976576 ram at 1005mhz (2.2v).

pci-e +0.1v
fsb +0.1v
mch +0.1v

my question:

on the mobo, in the lower right hand corner, of course depending on where you're standing in the place where one connects the "front panel headers" like pwr cable, reset cable, case speaker, HD led, etc, there appear to be two different places for the green power LED in the front panel, one within the grouping of all the wires (on what i'd call the top left, see page 25 section 10 of the manual) and the other place just to the left of this at the edge of the board (see page 26, section 11 of the manual), which seems to call for the exact same cable.

which is the correct place to connect the green power led light cable?

or does it matter?

thx.

 

[BillParrish]

What a superbly detailed question!

They are exactly the same in function and you can use whichever one want as far as I know. You WIN ! You made me look at the manual

What the deal seems to be is one is labeled PWR LED and one is labeled MSG LED and if the manual can be believed they both operate the same way.

S0 On
S1 Blinking
S3/S4/S5 Off

The difference is one is 3 pin and the other is 2 pin. If you have a case with a 3 pin connector for the power/sleep led you use the three pin separate header. I have never used one but I believe some cases have a combination green/yellow led and the three pin connection will cause the LED to be green when the computer is in the ON state and then the same LED will be blinking yellow when in sleep. If your power LED has only 2 pins use the one with the other front panel connections. And if you case has 2 LEDs do what you think best cause I don't know, and would just try and see. Many times I have had to modify the LED connection/connector by moving a pin/wire over in the casing.



Dear Motherboard Manufacturers.

Next big trade show please get all together, get drunk or whatever, hire some boobs if you have too, and sit down and decide on a standardized Front Panel connector layout and remember to include a couple of extra pins. ITs ONLY BEEN 20+ FRAKING YEARS . And after you sober up, tell the case manuf about it too. Thanks, appreciate that. Oh and it would be cool to have extra pins for a drive activity LED for each SATA drive so you might need to invite Intel too. One connector to rule the Front Panel...

Bill.

PS Dont be worried about giving the CPU a little extra voltage as your testing continues, you might be amazed at what 1.375V would do.

 

[silverspade14]

The long and the short of it is, put the fan on the NB, $5 and couple of dabs of RTV silicone adhesive and you do not have to worry about it. There is no way to determine exactly what your issue may be as there is insufficient information. What memory do you use and what is the manuf suggested voltage ? What were the FSB and multiplier setting etc. etc. It seems the cpu Heatsink/fan is working fine so that is probally not the issue.

Tech Spec
Capacity 2GB (2 x 1GB)
Speed DDR2 800 (PC2 6400)
Cas Latency 4
Timing 4-4-4-12
Voltage 2.1V

FSB was set at 333 and the multiplier was at 2

 

[BillParrish]

Tech Spec
Capacity 2GB (2 x 1GB)
Speed DDR2 800 (PC2 6400)
Cas Latency 4
Timing 4-4-4-12
Voltage 2.1V

FSB was set at 333 and the multiplier was at 2

Thanks, hmm so hopefully in the memory voltage settings you have the voltage set to +.3V. ( This may cause the memory voltage in PC Health screen to change color or say FAIL instead of OK, ignore it).

You should be able to use a memory mulitplier of 2.4 or 2.5 , I cant remember exactly what is available on the DS3, and that will run your memory on spec an it should make the machine a little snappier. 333 x 2.4 = 800MHz rate speed or if 2.5 is available 333 x 2.5 = 832 MHz for a very mild OC on the ram. It should easily do the 832MHz at tight rated specs but use a memtest86+ boot disk to verify for a couple of full runs before booting up the operating system. .



Has the BSOD repeated itself ? Typically that is a driver or software issue, most of the time thermal problems casue the machine to just freeze or reboot.

 

[svenski]

I haven't contributed to this thread for a while and was about to sell the DS3 3.3 until I stumbled over the new F13o beta BIOS. My board suffered from the well documented cold-boot issues up to F12 BIOS as soon as my Q6600 went up to approx 3.6Ghz independent of the multiplier, voltage etc I had similar issues with my previous E6600. I know that this had nothing to do with the CPU being limited as I could clock it much higher after 'low clock' cold-boot and the turning up the FSB with clock generator. To my understanding this was due to the board failing to provide the necessary overclock immediately when the board was powered up.
F13o seems to have solved these problems and I can now cold-boot at 3.8Ghz without problems. I will be trying higher setting but my vcore is getting a bit high for an aircooled system so this wont get prime stable. I am more than happy tough with these results for a cheapish Q6600.

 

[Punisher77]

Ok, I just finished reading pretty much the entire thread. Great work Bill!

I've got a Gigabyte GA-P35-DS3L that's giving me double boots (cold boot problem?) and loses OC all the time. I'm also having a problem running my external HD which is running on eSATA (doesn't work in USB mode either).

First question, do you have any suggestions about how to get my external drive working?

Second question, what do you recommend I buy after I sell this piece of sh*t MB?

Thanks!

 

[BillParrish]

Ok, I just finished reading pretty much the entire thread. Great work Bill!

Thank you.

I've got a Gigabyte GA-P35-DS3L that's giving me double boots (cold boot problem?) and loses OC all the time. I'm also having a problem running my external HD which is running on eSATA (doesn't work in USB mode either).

First question, do you have any suggestions about how to get my external drive working?

You do not provide sufficient details. Is the drive always plugged in ? Is it seen in the bios but not the OS ? Not ever seen ? etc. etc. This thread has some good ideas. http://episteme.arstechnica.com/eve/forums/a/tpc/f/24609792/m/644008649831 and in the thread it indicates the fault might be with Intel's Storage manager software. The OS should have been installed with AHCI drivers either with the floppy method for XP or using Vista built in. Set the bios to AHCI and raid even if you dont use that functions so the AHCI drivers that use "Hot Swap" are available to the system. Set the MB bios like you were using raid but just leave the drives as "standalone drive" in the raid setup.

Second question, what do you recommend I buy after I sell this piece of sh*t MB?

A Dell.

 

[Punisher77]

You do not provide sufficient details.

The drive is a 250GB 7200RPM Seagate (SATA1) in a Vantec Nexstar 3 connected via eSATA. When the drive is connected the bios just hangs in the HD detection phase and then eventually continues without detecting the drive. It randomly works and I've had it work by connecting it when in Windows (I'm running Vista) and then rebooting the machine. It's very random. I get the same issues when running on USB. This drive and enclosure worked fine with my old system.

 

[Punisher77]

I tried disabling legacy USB devices and my machine booted with the external drive this morning. I'll try it again tonight to see if it was a fluke.

I tried enabling AHCI but Vista just crashed. I'm guessing I have to enable the Vista drivers in the registry before enabling it. I'll try that tonight too.

 

[Sworkhard]

I am suddenly having issues with this board and rebooting. It appears to be shutting down but then the monitor goes black and the system lights stay on. I then turn off the power supply switch and when I turn it back on, the computer doesn't turn on when I hit the power switch. If I walk away from it for 10-15 minutes, I will find it turned on already.

Here are my specs:
Vista Home Premium 32-bit
E6420
DS3 rev 3.3 Bios F12
XFX Geforce 8800GTS 320MB
Patriot Extreme 800 1GBx2

I had the system overclocked at 8x400 for awhile, but then when I had issues I removed the overclock and still get the same problem.

I read about warm boot issues and tried my PCI-e at 101 and 105 Mhz, added +.1 volts to everything for stability, but still have the issue.

I also recently switched the HDD from the purple port to the yellow and disabled the AHCI options in the BIOS. Could this have something to do with it?

Sounds like the power supply, but it could be drivers. I had an issue like this back in the Athlon XP days and it was due to a PSU going bad.

 

[Sworkhard]

I tried disabling legacy USB devices and my machine booted with the external drive this morning. I'll try it again tonight to see if it was a fluke.

I tried enabling AHCI but Vista just crashed. I'm guessing I have to enable the Vista drivers in the registry before enabling it. I'll try that tonight too.

Yah, just enable the drivers in the registry, restart with the AHCI enabled in the bios, and install the latest intel chipset drivers to get the intel AHCI driver. I just finished doing this in vista 2 weeks ago with no problems.

 

[Punisher77]

Following this guide:

http://episteme.arstechnica.com/eve/forums/a/tpc/f/24609792/m/644008649831

Fixed my problems with my eSATA drive. Now it's time to try to fix the double boot issue when OCing.

 

[DerComissar]

What a superbly detailed question!

They are exactly the same in function and you can use whichever one want as far as I know. You WIN ! You made me look at the manual

What the deal seems to be is one is labeled PWR LED and one is labeled MSG LED and if the manual can be believed they both operate the same way.

S0 On
S1 Blinking
S3/S4/S5 Off

The difference is one is 3 pin and the other is 2 pin. If you have a case with a 3 pin connector for the power/sleep led you use the three pin separate header. I have never used one but I believe some cases have a combination green/yellow led and the three pin connection will cause the LED to be green when the computer is in the ON state and then the same LED will be blinking yellow when in sleep. If your power LED has only 2 pins use the one with the other front panel connections. And if you case has 2 LEDs do what you think best cause I don't know, and would just try and see. Many times I have had to modify the LED connection/connector by moving a pin/wire over in the casing.



Dear Motherboard Manufacturers.

Next big trade show please get all together, get drunk or whatever, hire some boobs if you have too, and sit down and decide on a standardized Front Panel connector layout and remember to include a couple of extra pins. ITs ONLY BEEN 20+ FRAKING YEARS . And after you sober up, tell the case manuf about it too. Thanks, appreciate that. Oh and it would be cool to have extra pins for a drive activity LED for each SATA drive so you might need to invite Intel too. One connector to rule the Front Panel...

Bill.

PS Dont be worried about giving the CPU a little extra voltage as your testing continues, you might be amazed at what 1.375V would do.

Agree 100%

 

[Punisher77]

Well, I tried installing the Gigabyte XPressRecovery 2 bios back up tool and it killed my disk partition. It now shows up as RAW instead of NTFS and I'm searching for a way to fix it. So far I haven't had any luck.