E2180, 3.6Ghz (1600FSB), possible to do on air?

Associate
Joined
14 Oct 2007
Posts
714
Location
Manchester
HI there. To put it briefly, I've currently got an E2180 running at 3.5Ghz on my P35C_DS3R, 2GB Corsair XMS2 memory, and an Arctic Cooling freezer 7 pro keeping the CPU itself nice and cold. Temps I get are about 30C normal use, and about 56C when under Orthos.

Now I was just wondering, do you guys think if I put a teeny bit more volts through it (currently at 1.45v), I'd be able to up the FSB to 400Mhz and drop the multiplier to 9x still using air, and obviously still have it stable?

I know it's not that much of a performance increase but I mean an extra 50Mhz (200Mhz effectively) on the FSB can't hurt anybody, and it'd also mean that I wouldn't have to run my (800Mhz) RAM faster than stock (currently at 840Mhz on a 2.4x mem multiplier).

And it also means I might get a few more marks out of my 3D Mark 06 score :D

Thanks guys :)
 
Your load temp seems fine so cant see any problem giving it a little more voltage and seeing what it can do. You may have better luck at 360x10. If you can do that, you may want to drop the multi and go for a higher FSB, though you may need more NB vcore or make some other BIOS changes to get it stable.
 
I've just tried 360x10 on one of my systems and it booted into Windows with just a little extra Vcore, so it's definitely possible. Why do you want to run 400x9?
 
Well you see I was planning on dropping the multi to 9 anyway because I don't think it's worth it getting an extra 100Mhz out of it when the FSB would only rise by 10Mhz. Not that I'll see a massive rise in performance at all, but I reckon getting another 50Mhz out of the FSB would be more beneficial to me than getting another 100Mhz out of the CPU.

I hear what you're saying though, I bet it would be more stable if I didn't raise the FSB speed as much

Edit:

Why do you want to run 400x9?

Because I keep hearing and I guess it makes sense, that a CPU with a faster FSB speed will be generally quicker than an even faster CPU with a slower FSB speed
 
I've just tried 360x10 on one of my systems and it booted into Windows with just a little extra Vcore, so it's definitely possible. Why do you want to run 400x9?
So his RAM will be at stock I guess... 360x10 means it'd be at 864 Mhz due to a 2.4x ratio, rathen than 400x9 where he could run 1:1.
 
Well you see I was planning on dropping the multi to 9 anyway because I don't think it's worth it getting an extra 100Mhz out of it when the FSB would only rise by 10Mhz. Not that I'll see a massive rise in performance at all, but I reckon getting another 50Mhz out of the FSB would be more beneficial to me than getting another 100Mhz out of the CPU.

Yes, I see where you're coming from - I had my head in "max overclock and hang the rest" mode.

It's definitely feasible either way, but 360x10 will run cooler as the clock cycles are lower.
 
I had over 3.7Ghz with a PA120.1, 3.6ghz is doable on good air I guess. The 10 multi could be easier to achieve it with.
 
*Update*: Tried it with 400FSB on a 9X multiplier, wouldn't have any of it. I tried giving it 1.5V first and the comp pretty much restarted itself because of the heat (I could hear the Freezer 7 go flat out before it restarted), this is all during Orthos. So I tried dropping the voltage to 1.475 and then it failed Orthos. Tried Orthos again and the comp restarted itself.

I've pretty much come to the conclusion that the only way you can achieve anything over a 360Mhz FSB with this cpu is to have it under water, or use a super dooper air cooler (:p)
 
Also - check for Vdroop

I use PC Wizard, but other monitoring programs also work (not CPUz) - run the system up at idle, and look at the CPU VCore, then run Orthos and recheck the CPU VCore. If it drops significantly then you may need even more VCore to stay stable.
 
there looks like there's some scope for upping the volts in your temperatures, just give it a go =D
 
Is it the CPU or the chipset that's dying from the heat?

I'm really not sure, how would you check? The only thing that happened was the computer just shut itself off (probably a safety thing) and restarted again.

Also - check for Vdroop

Ok I used pc Wizard and I think it dropped from about 1.44V idle to 1.41V on load, bearing in mind that at this current (stable) configuration I've got the voltage set at about 1.45V in the BIOS.

I guess this means that like Ben M said, I could probably up tha voltages a bit more to see if it works any better.

The thing is though, if it is the CPU that's causing everything to restart due to overheating, then wouldn't higher voltages just give me higher temps and hence just make it even more unstable?

I will try it when I get the chance anyway, although I doubt I'll be able to achieve a 1600FSB on this cpu without either a really well cooled tuniq tower or a water cooling setup
 
I don't think the CPU is overheating, I think it's dying from a lack of volts. There's really only one way to find out - feed it something outrageous like 1.5V and see what the idle temperatures are like. If it is dying from droop then the idle temperatures will be high because of all the voltage you're feeding the CPU, but the load temperatures won't be nearly so bad as the motherboard won't be giving it all the volts under load.

As far as the overheating check on the chipset - if you cannot comfortably hold the northbridge cooler, it's probably a bit on the warm side.
 
Right thanks WJA96, I will try that. I think I read somewhere that people generally don't give the cpu any more than 1.55V, so I guess I should be ok with 1.5V.

And about the Vdroop thing, is that normal for a setup like this or is it down to poor quality components in the mobo or PSU or even both?
 
And about the Vdroop thing, is that normal for a setup like this or is it down to poor quality components in the mobo or PSU or even both?

Sadly, it seems to be more and more common, even in the very best motherboards these days. It's an ASUS speciality though.
 
Back
Top Bottom