Clearly I'm not an overclocker

Liver

Supreme [H]ardness
Joined
Oct 24, 2005
Messages
6,368
Never been able to do it right. Had issues, and really don't know what I am doing wrong. Other the flip side, I've half assed tried to learn, but never really dived into it.

So a couple of questions. I have the Intel Bad Axe 2 mother board (975XBX2) with a q6600 quad core processor, coupled with some nice Corsair memory (Dominator 6400 in 4gb [2x2] variety).

Zalman 9700 air cooling with Artic Silver between the two.

I followed an online guide on how to overclock the chip and I got from 2.4 to 2.5 :eek:

I slowly increased the FSB until it would not post. Never made it into Windows. That is all I changed. No voltage increases, nothing. Kept everything automatic to limit the variables (followed the guide on Tom's).

http://www.tomshardware.com/forum/240001-29-howto-overclock-quads-duals-guide

I can totally understand if I got a non overclocking chip, but I think I can get better than a 4.1% improvement.

I am convinced I did something wrong, or neglected a step. Any help would be great.

I'm wondering if just by increasing the FSB that the memory was unable to keep up.

If that is not the guide to follow, please help a noob out!

I don't care if it frys the system, as long as my SSD makes it out.

edit: And I clearly put this in the wrong forum.
 
Last edited:
Did you run the memory at 1:1? Take the ram out the picture until your FSB is stable. Once that's done, start working to get the ram running at the highest possible frequency with the best possible timings at the lowest possible vDIMM.
 
I think I did, but I will check that. I really did try to follow that article.

I left all the memory timings on automatic. Then I changed the settings to what the memory said (i.e. 4-4-4-12 I think). Took it right off the memory (as stated by the article)
 
You want to make sure your divider is set to 1:1, that way you won't hit pc6400 speeds until you're at 400 fsb.
Posted via [H] Mobile Device
 
edit
I know the q6600 chip is kinda dated, so the google searches I made clearly pointed to the Tom's article.

Is there a better article / guide I could follow? I'd like to try again before I gave up on overclocking this chip. Quite honestly I'd like to overclock to delay upgrading as I don't think I really need to buy a new computer (yet). I think I may be able to get through 2010 with this.

You want to make sure your divider is set to 1:1, that way you won't hit pc6400 speeds until you're at 400 fsb.
Posted via [H] Mobile Device

Will do. I have a feeling I already did that, but it is easily possible I made a mistake.
 
Q6600 is somewhat dated but is still a very fast chip. Depending on what you're using it for, it should be more than enough horsepower, even without overclocking. OC'ing takes practice and patience. You should be able to easily get it to 3.2GHz without increasing any voltage. Make sure the RAM timings are set to auto and increase CPU, VTT voltage. You can safely increase to over 1.4V without worry.
 
memory is the most important when overclocking. As you up the FSB so does your memory. make sure you have set your memory divider that corresponds with your memory speed capabilities.
 
memory is the most important when overclocking. As you up the FSB so does your memory. make sure you have set your memory divider that corresponds with your memory speed capabilities.

Definitely. I always start by under-clocking the ram to take it out of the mix.
 
So I did not underclock the memory nor did I run it 1:1. By looking at the BIOS I really can't figure out how to do so.

I did input the settings for the memory that are listed on the memory (5-5-5-18).

I am going to search for recommendations specifically to the 975XBX2 motherboard, but if you already know how, please let me know.
 
So I did not underclock the memory nor did I run it 1:1. By looking at the BIOS I really can't figure out how to do so.

I did input the settings for the memory that are listed on the memory (5-5-5-18).

I am going to search for recommendations specifically to the 975XBX2 motherboard, but if you already know how, please let me know.

Turn the memory timings to Auto and try it again. The memory speed may not jive with the inputed timings.
 
5-5-5-18 is pretty loose. You should get a little OC with timings like that even with stock voltage.
 
Alright I need some more education, so I need to ask this bone head question.

My FSB is 1066 (on the motherboard). When I increase the FSB for the chip (266 to 275) the mainboard goes to 1120 and craps out.

I am sure I am not doing something right. How do I rectify it? I am making a basic mistake and don't know where to go from there.

I appreciate all help.
 
Alright, this is based off of screenshots I found online, so who knows how accurate they are with newer bios versions (I'd probably update the bios also while you're at it so you have the latest one).

Under CPU Voltage Override, leave it on AUTO, but if you have stability issues change it to 1.3 to 1.35V or so. I don't remember off the top of my head what the VID is for Q6XXX, so you might be able to go up to 1.375V or more depending on your cooling. Ideally you want this as low as possible.

FSB Frequency Override - Put it on Manual

FSB Frequency - Type in a number...266 is stock. A Q6600 should run 333 without too much trouble, especially with the voltage bump above.

SDRAM Control - Memory Frequency - put it at 400mhz...that should take your memory out of the OC equation to start. You can move it up to 533 or whatever after you figure out what your max OC is on your CPU.
 
Fairly sure part of the problem here is the older chipset of the BadAxe2. The 975X isn't going to OC nearly as well with a quad core (even a 65nm quad like the Q6600) as a P35 or P45 based board would.
 
Fairly sure part of the problem here is the older chipset of the BadAxe2. The 975X isn't going to OC nearly as well with a quad core (even a 65nm quad like the Q6600) as a P35 or P45 based board would.

I know the P5W DH will run the 45nm quads and it's running on the same 975 chipset with a proper bios update. The BX2 and the P5W were two of the best boards out back during that time frame. I figure if it will run the 45nm quads at 1333mhz FSB, it should be able to run the 1066mhz bus CPU's at 1333mhz which would be 3Ghz on the Q6600.
 
I know the P5W DH will run the 45nm quads and it's running on the same 975 chipset with a proper bios update. The BX2 and the P5W were two of the best boards out back during that time frame. I figure if it will run the 45nm quads at 1333mhz FSB, it should be able to run the 1066mhz bus CPU's at 1333mhz which would be 3Ghz on the Q6600.

Makes sense, but my experience with the 680i and quads doesn't go the same way. I could get my E6420 near 3.8Ghz on my first 680i, but it would hard lock above ~2.8, 2.9Ghz with the X3220 (like a Q6600). The Abit IN9-32X-MAX I got to replace it would run 45nm quads with a BIOS update (unlike the original reference 680i) but it didn't OC the X3220 any better. I never even managed 3Ghz with it.

Bought a 780i and did a bit better -- 3.2-3.4Ghz range. My current P45 can get it up near 4Ghz but the vcore is too high for my comfort; this isn't a great OCing chip by any stretch, but the boards I've been through were definitely the limiting factor.

The BadAxe2 definitely had a reputation as a great OCer, though, so there's potential. I don't have any first hand experience with it so I'll leave that to the Intel enthusiasts. Good luck OP.
 
Did you run the memory at 1:1? Take the ram out the picture until your FSB is stable. Once that's done, start working to get the ram running at the highest possible frequency with the best possible timings at the lowest possible vDIMM.

This, it could be pushing your ram too hard as I made this mistake the first time I started overclocking. I left the ram in locked ratio and since it is involved with the FSB too, it wouldnt post.

I just went ahead and unlocked my ram, it auto adjusts the ratio to run the ram at stock speed.
 
Never been able to do it right. Had issues, and really don't know what I am doing wrong. Other the flip side, I've half assed tried to learn, but never really dived into it.

No voltage increases, nothing.

I can totally understand if I got a non overclocking chip, but I think I can get better than a 4.1% improvement.

I am convinced I did something wrong, or neglected a step. Any help would be great.

You didn't increase your voltage and you expected a good overclock?

I think you've answered your own question. Maybe you aren't an overclocker because you didn't increase voltage. The thing you never learned is the fact that you almost always need to increase voltage. The thing you did wrong and the step you neglected was not increasing voltage. Am I clear yet? Voltage voltage voltage! :p That's a core tenet of overclocking! :)

The Q6600 can easily go to 3.2-3.6GHz (and higher) with 1.3V to 1.5V.

The same goes for your ram. If your FSB isn't getting properly divided for the memory or you are pushing slightly higher numbers than it is rated for at certain timings, you should also look at increasing memory voltage.

The first thing I did when I got my Q6600 was to throw on the biggest air cooling I could find, bump the voltage up as high as I was comfortable with, set the memory divider as high as I could so the memory would never get in the way, and then let the FSB fly. I generally go for the highest I think I can get first, and then I move backward until I find stability. I started with 3.8GHz on my first attempt because I like that number. I don't start low and move up 1mhz FSB at a time. I think that's a total time waster. I'd rather know my ceiling first and work backwards. You can't cause any damage unless you are applying more voltage than the chip can take (Intel spec for this chip is commonly accepted at a maxmium of around 1.5V) and if you don't have properly cooling. The worst that can happen is failure to post. Once you have your CPU at a stable speed that you are happy with, then it's time to work on the memory. Start reducing the divider and pushing tighter timings. I can squeeze over 1260MHz out of 1066MHz rated ram (again however, voltage, voltage, voltage!)
 
Last edited:
Makes sense, but my experience with the 680i and quads doesn't go the same way. I could get my E6420 near 3.8Ghz on my first 680i, but it would hard lock above ~2.8, 2.9Ghz with the X3220 (like a Q6600). The Abit IN9-32X-MAX I got to replace it would run 45nm quads with a BIOS update (unlike the original reference 680i) but it didn't OC the X3220 any better. I never even managed 3Ghz with it.

Bought a 780i and did a bit better -- 3.2-3.4Ghz range. My current P45 can get it up near 4Ghz but the vcore is too high for my comfort; this isn't a great OCing chip by any stretch, but the boards I've been through were definitely the limiting factor.

The BadAxe2 definitely had a reputation as a great OCer, though, so there's potential. I don't have any first hand experience with it so I'll leave that to the Intel enthusiasts. Good luck OP.

Yeah, I never played around with the 680i boards or the BadAxe2, but I did have the P5W. Great board for its time. My father still runs it with my old E6600. I think it's about time for a quad core upgrade for him :D.
 
Aren't Intel boards notoriously bad to overclock on? They're not really made with overclocking in mind, more for stability. That's the impression I got from reading around.
 
Aren't Intel boards notoriously bad to overclock on? They're not really made with overclocking in mind, more for stability. That's the impression I got from reading around.

Not "notoriously bad" but they are more for stability than overclocking. That being said, if you've got a decent chip I see no reason you couldn't get a pretty good overclock even on an Intel board.
 
The bad axe boards had overclocking in mind. They are some of the best 975X boards out there along with the P5W DH.
 
Back
Top