I've heard that DFI boards are legendary overclockers, and with the limited 8.0x multiplier on the Q9450, I'm going to need a great overclocking motherboard to pump some real power into it. In a perfect world, I'd love to hit 500 FSB and 8.0x multiplier making a frequency of 4.0 GHz. I may not be able to make it that far, but if I could get to 3.6 or so, I'd probably be happy.
Does anybody know enough about this motherboard to help me out?
Also, my RAM is 4 GB (2 x 2 GB) of G.Skill DDR2 @ 1000 MHz, and my CPU cooler is the Xigmatek HDT-S1283 120mm Rifle CPU Cooler.
While DFI makes a great motherboard, there are a lot of other factors that determine the top speed of the CPU. At this time, most people that I've read about are reaching top speeds in the area of 3.2-3.6ghz with the Q9450. In fact, I've only read of one person so far that reached a speed of 3.6ghz, though there are likely others. Heat is a major problem, and not only for the CPU, but for other components. The northbridge, for instance, can be particularly sensitive to heat in many motherboards. To hit high speeds, you also have to raise voltages considerably and that can lead to instabilities, as well as higher temperatures. Cooling because an ever increasing problem as higher clocks, with their attending higher voltages, are attempted.
To try to give an example. I have hit as high as 4050mhz with my QX9650, but I could maintain stability while running Prime95. The highest stable clock I've been able to hold so far has been 3780mhz (9x420). I might be able to get it a bit higher with further experimentation, but its not a sure thing. Every CPU is different, and while one CPU may hit a high overclock, another one of the same type may only get a small overclock. Keep in mind that overclocks of any chip are not guaranteed. They are very much governed by sheer luck in getting a good CPU that will take a high overclock.
So I could get stuck with one that could only OC to 3.2 GHz? Hmm. =\ Makes me wonder if I should just pick up a Q6600 instead, since they're pretty much guaranteed to hit 3.6 GHz nowadays with the G0 stepping. Does this sound like a good idea?
Alternatively, you could get a Q9550. But keep in mind what I wrote about guarantees, there are none. Even though a lot of people may report getting 3.6ghz from a Q6600, you don't hear about those who tried for high overclocks and fail.
O well... Thats life. Sailer you shouldn't put him off getting that CPU because it may not overclock. You took the same risk when you brought yours and it's a risk every overclocker takes everytime he/she purchases a new processor.
nman729, I have been reading through the threads looking for some example overclocking using a rampage formula and a Q9450 as I just picked one up from Tokyo yesterday and will be overclocking it as hard as I can.
Would you mind listing some voltages, multipliers etc I am collecting info for when I overclock mine. Thanks
I think northbridge heat will probably be the biggest problem in terms of heat. I have already oc'd mine but my mobo or cpu (don't know which one) is holding me back. I cannot load Windows with anything past 445 FSB and that's not stable at all. Oc'ing with 8X multi is very challenging. And I have upped my Vcore to 1.5 and I could not get past 445x8. However, Oc'ers should not despair, because it's fairly easy to Oc to 3.2 GHz and 3.4 GHz (provided you have good cooling). Anything else is unnecessary and will probably ruin your mobo/cpu if you keep pushing voltages up. And don't worry about performance, this is the best Yorkfield cpu out there for the money.
Fellows its not your Northbridge giving you fits as even the old P965 chipset would easily give 450ish FSB on a good name brand M/B. It most likely the cpu is just FSB sensitive. I have a Q6700 that gives up after 370 FSB with the multiplier locked at 6x. That's even slower than stock speed! A 48 or 45 intel chipset board should give a stable 500+ FSB if you get a processor that will tolerate it. Start out with a proper OC first see what FSB it will handle using a locked 6x mutiplier first.