• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

***The Official Yorkfield Q9300/Q9450/Q9550 Overclocking Thread ***

Hesky I have run BLEND now for 2hrs 10mins, would this indicate stability in your opinion.
I am gonna leave it running for rest of day while I am out
 
Two hours blend is a good test of overall system stability, you have ran it long enough to have at least run a good 30mins of 8k/10k small ffts (blend is both small and large ffts lengths) which would have really pushed your cpu. I would say it’s pretty damn stable now :)
 
ATLANT3AN said:
What truly is hands on heart the safest cpu voltage for these Q9series? which wont result in any loss of life span? I dont know If I should go back to 3400 with 1.2v. I cant change computer for 3 years and when I get a new one this will become my second machine, so really 5 years is the lifespan of my machine, if It will easily reach that then I dont mind
Click to expand...

ATLANT3AN, 1.3v and your current temps are truly safe for 24/7 running, its only a smidge over stock voltage and well within Intel’s specified limits for yorkie’s, which are guidelines for a specific max voltage on much much less effective cooling than what you are currently running. Your cpu will live long I swear on it :)
 
One tiny notch up on the vcore won’t hurt :) your unlikely to run your cpu at those load temps again, and from what you said before your temps are fine running prime anyway.

I’d just do that and not bother with anymore stability testing and just get on enjoying your new pc. :)
 
Thanks Hesky82 for your support, I have upped the cpu volts to 1.320v, although that is what it says in CPU-Z, It is set slightly higher in the BIOS. I did this because PRIME95 ran again today one core failed at 1 hours 16mins, So I thought I would up the volts buy 2 levels and then see if I can get it to run 5 hours by which point I will not mess with this again.

So 1.320v (3520MHZ) for 5/6 years lifespan at least
 
Last edited:
Yup that’s just vdrop for you, quads suffer from it the most even on top notch boards with good cpu power regulation.

I've found that CPU-Z's is usually pretty accurate, comparing it to the reading I get from a DMM connected to the vcore measuring points on past and present mb's they tend to be near identical.
 
If your using applications that can utilise the new SSE4.1 set of instructions then I think its worth the extra dosh.

Then again nehalem is just around the corner which will enable all 54 SSE4 multi data extensions with the updated SSE4.2 version, giving it a full SSE4 complement.

Nehalem’s new microarchitecture maybe something worth saving up for when applications start using these rather new instructions.
 
Thats a nice speed, 3.7ghz, that gotta be easy equal to a 4ghz 6600
I am very tempted to sell my Q9300 now I have had a play with it and get a Q9450
 
Hi guys, my forum account has finally been validated -- after just about a week ! ;-)

I have a GA-EP35C-DS3R board, a Q9300, 2GB of 5300 memory, and a Akasa AK-965 cooler.
I run it easily at 3Ghz using the 'auto' setting, which pushes the voltage to 1.3V. The memory runs at 800mhz do far too.

One puzzle tho: The 'die temperature' is very, very low (14, 15C !) with the fan at low speed (750 rpm) but the cores are pretty hot (42,42,46,41) at idle; one of them is constantly about 4/5C highter than the others.
The puzzle is that at stock speed I get pretty much /exactly/ the same temperatures, within maybe one degree. Oh and with busy cores I get about 55+

The extra data is that I had to hand patch my linux kernel to recognize the 0x17 CPU family and read the temperatures; do I'm not entirely certain about the actuals values. I'm about 90% confident only :-)

+ So what sort of temperatures do you guys get on these 9300 ?
+ What explains such a use difference between the BIOS "die" temp and the cores ?
+ Should I lap it to try to get even temps on everyone ?
 
OK I lapped the Q9300, coated it with a very thin layer of Silver 5, and it made a total of zero difference. None. Nada, Nitch :-) Even the CPU3 that is hotter than the others, same. Wierd!
 
I was amazed at how uneven the base of my Akasa cooler was, but I recon the top of the Q9450 isn't flat either. Maybe the Cooler deforms the heat spreader on the CPU and makes even conntact that way.

I did give it a good dollop of AS5 though.
 
BusError said:
OK I lapped the Q9300, coated it with a very thin layer of Silver 5, and it made a total of zero difference. None. Nada, Nitch :-) Even the CPU3 that is hotter than the others, same. Wierd!
Click to expand...

I have the lapping kit in front of me and am thinking shall I, shant I
suprised it made no difference, was the heatspreader flat to begin with.
 
@bikes: I didn't. I use that Alaska and I can't take it apart to access the flat base sufficiently to make sure I polish it flat.
@Banny: The top of the Q9300 was not entirely flat, wich had me think it would work better. I added a 'grain of rice' of S5 and spreaded it evenly with a plastic spatula.

But I'm still surprised at the difference between 'die' and 'cores' temperatures...

@BinBoy: are you overclocked at all ?

BTW I switched to manual voltages and went down from 1.3V to 1.25V, still at 3Ghz and still perfectly stable.
 
You must log in or register to reply here.
Back
Top Bottom