Q9550 OC + GA-P35-DS3L = Headache

[HigherGround]

I scored a cheap Q9550 (Microcenter) and decided to replace my 2 year old Q6600. Just to be sure that the board can handle higher bus speeds I run my Q6600 @ 400Mhz FSB for the last few days. Today I dropped in Q9550, booted straight to BIOS and loaded defaults (333x8.5). Saved and booted to Win 7. Run prime for about an hour with no errors. Temps were pretty good (high 20s/low 30s idle, low 40s loaded). Board vdrooped the Vcore to 1.09V during stress test and 1.11 when idle. Encouraged by the temps I bumped the FSB to 400 and set multiplier to 8 (3.2Ghz), again with everything set to Auto ... I got greeted with a massive BSOD. I dropped the FSB to 375 (3.0Ghz), again with no luck (this time windows locked up). I tried bumping all voltages, again with no success. Finally I tried 400*6 (2.4Ghz, which is a 400Mhz underclock) and again it locked up. It seems that I cannot go over 333Mhz FSB with this chip, although it did run @ 400Mhz with my Q6600. Oh, and the memory was locked at or below 800Mhz during all tries with very relaxed timings (6-6-6-18s).

Can anyone help?

System specs:

Q9550 (VID = 1.15V)
Cooler Master H2O
GA-P35-DS3L Rev. 2.0 BIOS: F9 (Latest)
4GB OCZ Titanium 800Mhz @ 4-4-4-15
Corsair HX520
8800GT (Stock voltages)
3x HDD
5x12cm fans (all @ ~7V)

 

[BlahBlahYouToo]

what stepping is that 9550?
my E0 does 400 x 8.5 all day, undervolted, no sweat.

 

[Indus]

I scored a cheap Q9550 (Microcenter) and decided to replace my 2 year old Q6600. Just to be sure that the board can handle higher bus speeds I run my Q6600 @ 400Mhz FSB for the last few days. Today I dropped in Q9550, booted straight to BIOS and loaded defaults (333x8.5). Saved and booted to Win 7. Run prime for about an hour with no errors. Temps were pretty good (high 20s/low 30s idle, low 40s loaded). Board vdrooped the Vcore to 1.09V during stress test and 1.11 when idle. Encouraged by the temps I bumped the FSB to 400 and set multiplier to 8 (3.2Ghz), again with everything set to Auto ... I got greeted with a massive BSOD. I dropped the FSB to 375 (3.0Ghz), again with no luck (this time windows locked up). I tried bumping all voltages, again with no success. Finally I tried 400*6 (2.4Ghz, which is a 400Mhz underclock) and again it locked up. It seems that I cannot go over 333Mhz FSB with this chip, although it did run @ 400Mhz with my Q6600. Oh, and the memory was locked at or below 800Mhz during all tries with very relaxed timings (6-6-6-18s).

Can anyone help?

System specs:

Q9550 (VID = 1.15V)
Cooler Master H2O
GA-P35-DS3L Rev. 2.0 BIOS: F9 (Latest)
4GB OCZ Titanium 800Mhz @ 4-4-4-15
Corsair HX520
8800GT (Stock voltages)
3x HDD
5x12cm fans (all @ ~7V)

I have that exact motherboard and was pondering that exact CPU upgrade. I have an E6550 oc'ed so you can try my settings and maybe get somewhere:

Use F8 bios
Graphics Booster: AUTO
CPU Clock Ratio: 8.5
Host Clock Control: Enabled
CPU Host Freq: 400
PCIExpress Frequency: AUTO
CIA 2: DISABLED
Performance Enhance: STANDARD
System Memory Multiplier: 2.0

System Voltage Control: MANUAL
DDR2 Overvoltage Control: +0.2V
PCI Express Overvoltage Control: Normal
FSB Overvoltage Control: +0.1V
(G)MCH Overvoltage Control: Normal
CPU Voltage Control: 1.35V to start (if system is stable, lower it.. mine runs a 600 mhz overclock while undervolting at 1.15)

Let me know how this goes, I'm very curious since I might be upgrading to your cpu too.

P.S. I have heard F9 bios is buggy and doesn't play nice with Q6550 but F8 does.

 

[RussianSensation]

1. Lock PCI Express to 100
2. Did you try increasing VCore to say 1.30V?
3. Did you try increasing your FSB termination voltage to 0.1V/0.2V and MCH voltage by 0.1V?

You should have posted on here first before trying a Q9550 on that DS3L P35 board. Other than Abit P35 boards, most P35 boards (esp. DS3L P35 with its 4-phase power regulation) are going to disappoint for Q9550 overclocking. If you were already running Q6600 @ 3.6ghz, I hate to tell you but your Q9550 won't be an upgrade.

You should have sold your Q6600 + Mobo + Ram and got a Core i7 system. I had the same mobo + cpu as you and that's what I did. It was cheap to upgrade.

 

[faxon]

I would say the board is gonna be your problm more than anything considering that most of them were poor quad oc boards and they didn't do as high fab as a good p45 board can do now

 

[HigherGround]

Thanks for the answers. It looks like the board doesn't like anything above 370 FSB. Tried 400x7, which is an underclock for the CPU and it froze during startup. I set both FSB and MCH voltages to max values (+0.3V), but again, same result - hard freeze during bootup.

what stepping is that 9550?

it's E0. It runs @ 1.08V loaded @ stock speed. I'm sure this chip has more to give, but I'm beginning to doubt this board will do the job.

Update: Decided to start dropping Vcore at stock speeds. Now I'm @ 1.05V, did a 20 min BurnTest and looks solid.

Another Update: I'm at 1.025V now and it seems to be working, temps are great (45C for hottest and 39C for coldest core under full load).

 

[Indus]

It could be the cpu too. I mean you're never guaranteed an overclock but some do generally perform very well. For example another person who has the same board as mine cant even hit 2.8 ghz while I have hit 7 x 500 once.

 

[HigherGround]

It could be the cpu too. I mean you're never guaranteed an overclock but some do generally perform very well. For example another person who has the same board as mine cant even hit 2.8 ghz while I have hit 7 x 500 once.

dunno, i'm thinking it's the board because any kind of FSB increase, even with lowered multiplier (therefore significantly underclocked) = hard freeze. I'm running this CPU @ stock speed at 0.98V atm and it looks solid.

 

[error8]

It's something that you're doing wrong. It's impossible for that board not to hit 400 mhz FSB with that chip.
Set everything to manual. Increase MCH voltage, put CPU voltage at 1.38 V, set ram timings manually, increase SB voltage, VTT voltage everything, but not to the maximum, 0.1 V or around that.

 

[n7]

EP45-UD3R in your future perhaps? :twisted:

DS3L is a garbage board for OCing quads, sorry.
Even the P35 DS3R isn't great for quads.

As others have mentioned, up MCH & VTT [FSB Overvoltage] voltage.

I imagine vdroop is quite bad as well on that mobo...i assume you did try a much higher vcore, like say 1.3-1.35 at least?

And absolutely make sure you're keeping the RAM @ the 2.00 ratio (1:1) w/ loose timings.

 

[HigherGround]

it's utter crap. I bumped all voltages (MCH and VTT +0.3, which is max), set FSB to 400, set ram at 1:1 with 6-6-6-18 settings, bumped CPU voltage to a healthy 1.3V and tried booting @ 2.4Ghz (6x multi). No go, system froze during windows logo animation.

This board vdroops by 0.05-0.07V depends on load.

EP45-UD3R in your future perhaps? :twisted:

DS3L is a garbage board for OCing quads, sorry.
Even the P35 DS3R isn't great for quads.

As others have mentioned, up MCH & VTT [FSB Overvoltage] voltage.

I imagine vdroop is quite bad as well on that mobo...i assume you did try a much higher vcore, like say 1.3-1.35 at least?

And absolutely make sure you're keeping the RAM @ the 2.00 ratio (1:1) w/ loose timings.

 

[amenx]

I have that board and was reluctant to go to a 45nm quad myself. ATM have an e8400 and it can easily do 500 FSB on it. But with quads its a whole different ballgame. Also I wouldnt want to OC a quad too high on it because as mentioned its a 4-phase power regulation and it may not handle it well.

 

[error8]

I would try to use some heatsinks on the mofsets and improve the airflow around that area. Maybe they become too hot to function properly, when you increase the FSB with that quad.
At least take a fan and point it towards that area.

 

[EarthwormJim]

Everyone seems to be ignoring the fact that you had a quad at 400mhz prior to this Q9550...

Have you tried something as simple as updating the bios?

 

[HigherGround]

Everyone seems to be ignoring the fact that you had a quad at 400mhz prior to this Q9550...

Have you tried something as simple as updating the bios?

That's true, Q6600 did run @ 400*6 on the same board with the same BIOS (which by the way is the latest). I had to disable USB HDD for it to work though. For some reason this mobo does not like 45nm quads. I'll try pointing my 120mm fan directly at NB and see if that makes any difference.

Update: The board craps out somewhere between 360 and 370Mhz FSB. Right I'm running at 3.06Ghz @ 1.05V and it looks good. 3.2Ghz is no go at any voltage (tried up to 1.35V).

 

[RussianSensation]

Although we did mention to bump FSB termination voltage by +0.3V, don't run this. 45nm quads crap out beyond 1.45V. I believe on P35 DS3L, Normal is 1.2V. So the max you will want to run is +0.2 FSB termination, but ideally +0.1V if that's enough for 370 FSB.

I was just trying to see if this makes a difference. However, running MCH and FSB termation at +0.3V 24/7 will kill this board and the processor as well.

 

[BababooeyHTJ]

Although we did mention to bump FSB termination voltage by +0.3V, don't run this. 45nm quads crap out beyond 1.45V. I believe on P35 DS3L, Normal is 1.2V. So the max you will want to run is +0.2 FSB termination, but ideally +0.1V if that's enough for 370 FSB.

I was just trying to see if this makes a difference. However, running MCH and FSB termation at +0.3V 24/7 will kill this board and the processor as well.

They do?

Although, I agree that I wouldn't be running my VTT in excess of 1.45 for 24/7 on air.

 

[RussianSensation]

They do?

Although, I agree that I wouldn't be running my VTT in excess of 1.45 for 24/7 on air.

Ya they do, even 1.45V is too much for 45nm Penryn. While guys at Xstremesystems are awesome benchmark overclockers, you have to take their findings with a grain of salt for 24/7 100% load operation.

http://www.anandtech.com/weblog/showpost.aspx?i=428

"(Editor- Raja in case the police get involved) made the mistake of using a very high VTT termination voltage of 1.51V (VTT is used to terminate data lines between the MCH and CPU).

We know users are running VTT voltages even higher than ours on 45nm processors and probably have not had a problem yet. We will run high VTT voltages in short bursts to test the limits of the board and CPU. However, this is the first time we have tried anything over 1.45V on a 24+-hour basis to test application stability.

Let this be a warning – do not go over 1.4V maximum for 24/7 use!

This is our first 45nm Quad core processor we managed to kill outright during testing. We hope it is the last one too. The problem is that we also have a Q9300 that is on life support after experiencing a 36-hour run at 435FSB with VTT set to 1.45V. While our experiences might not represent results elsewhere, we thought our advice to just, “Say no to high VTT"

Since P35 DS3L defaults at 1.2V FSB termination IIRC, running +0.2V is already maxing out that quad.

 

[Dadofamunky]

Looks like a new mobo is definitely in order, unless you can live with 3 GHz max on that chip (it can do a lot more than that). It's amazing it can run 45nm quads at all.

 

[mmsandi]

just try to flash to f8 bios

 

[Markfw]

EP45-UD3R in your future perhaps? :twisted:

DS3L is a garbage board for OCing quads, sorry.
Even the P35 DS3R isn't great for quads.

As others have mentioned, up MCH & VTT [FSB Overvoltage] voltage.

I imagine vdroop is quite bad as well on that mobo...i assume you did try a much higher vcore, like say 1.3-1.35 at least?

And absolutely make sure you're keeping the RAM @ the 2.00 ratio (1:1) w/ loose timings.

Hey, I had a Q6600@3.2@100%load for months (or years??) 400*8, no problems...

I thought it was a great board.

 

[HigherGround]

just try to flash to f8 bios

tried that as well. That board was a POS. I've just finished mounting this CPU in a ga-p45-extreme. I'm @ 3.6Ghz with room to spare (1.18Vcore).

 

[hclarkjr]

tried that as well. That board was a POS. I've just finished mounting this CPU in a ga-p45-extreme. I'm @ 3.6Ghz with room to spare (1.18Vcore).

i have a 6600 currently and have been kicking around getting a 9550 soon, since we have the same board and such could you post some benchmarks comparing the 2? have you noticed a big difference? i too have a P45 board i will be using for it when i buy it. thanx

 

[n7]

tried that as well. That board was a POS. I've just finished mounting this CPU in a ga-p45-extreme. I'm @ 3.6Ghz with room to spare (1.18Vcore).

Nice!

:thumbsup:

 

[HigherGround]

i have a 6600 currently and have been kicking around getting a 9550 soon, since we have the same board and such could you post some benchmarks comparing the 2? have you noticed a big difference? i too have a P45 board i will be using for it when i buy it. thanx

sorry, i didn't have time to benchmark my old q6600, but subjectively it definitely does feel a little bit snappier. I bought mine only because I was visiting my folks on Long Island (I live in Europe) and they have a Microcenter right there at Westbury. I'd already sold my q6600 for $180 so basically that was like a free upgrade