Quote:
Originally Posted by Nazu
around 58 C
Printable View
Quote:
Originally Posted by Nazu
around 58 C
I'm toying with a friends q6600 on my board and can't get it above 375 fsb. I had the same problem with the chip in my sig but thought it was my cpu. Could it be the board? SHouldnt' i be able to hit 400+ on this board?
680i isnt very good for quad fsb, P35 would be better.Quote:
Originally Posted by iadstudio
Very nice :clap:Quote:
Originally Posted by qujck
LOL check your mount before you blame the CPU.Quote:
Originally Posted by T3rm1
Good job,Quote:
Originally Posted by qujck
Have you tried small FFT ?
I did :(Quote:
Originally Posted by GripS
Looks ok to me. I have no clue why my cpu is so hot.
Btw: Im using Scythe Mugen cooler.
Infinity/Mugen is not the best out there. You could mod the mounting system and replace the push-pins with bolt+nut+spring mounting (...or just buy the TR s775 bolt thru kit).
Will lowering my multiplier to 8x result in significantly better performance with a FSB of 450 or more? I ordered new ram that will run at 1100Mhz+. Do I want the highest FSB possible? I'm confused. Right now I'm at 9x400.
heres mine. i tryed to get to 3.4 but it would crash after an hour with prime95. i dont want to increase my voltages too much so ( right now at 1.3vCore) i'm going to stay at 3.2.
Yes, the funny thing was that if i'm using only small fft's the former tries were stable everytime but if i did the whole test it was somestimes unstable.Quote:
Originally Posted by Devious
So I'm using the whole range of fft's to make sure that it is stable.
I was just using that cooler. It does quite well (even with push pin mounting) with 2x120mm fans in a push/pull config.Quote:
Originally Posted by T3rm1
Recheck your mount again. Use a good quality paste and try to push both pins down on opposing sides of the heatsink. Make sure you hear a faint 'click' on each pin before you call it good. Flip the board over and make sure all pins are locked in.
I installed the cooler when I had the mainboard infront of me. I checked the bottom side of the mainboard and all pushpins were ok.Quote:
Originally Posted by GripS
There are only 3 possibilities right now:
(1) Im using the silicon paste that was in the package (already ordered ArcticSilver 5 and Zalman paste). But even if the temperature is 4° cooler with the new paste, its still too hot.
(2) I didn't lap the cpu.
(3) The cpu has 1.3 VID. Everyone else with good temperatures have something around 1.28 or less.
I ordered a new cpu. This time the boxed version. Hope Im lucky with it.
Looking better & better.. needs a bit more testing though
http://aycu09.webshots.com/image/338...4893642_rs.jpg
holy crap that quad OWNZ!
I'm in the same position. Have 400*9 with Ballistix 8500 running at 3:4. I've been trying to get 450*8 but am struggling to get it stable (although I'm confident I can achieve it when I find the time to spend on it). I did notice that the temps dropped a bit when running 450*8.Quote:
Originally Posted by Hans Gruber
Would someone be so kind as to detail the benefits of running a lower multi with a higher FSB to us 'noobs' :p:
Quote:
Originally Posted by ~aoe~
Thanks for replying to my question. What's the point of the Q6600 being a 1066mhz processor when my motherboard supports 1333mhz? I'm running at 3.6Ghz 9*400 which means my FSB is 1600Mhz, well above the 1333Mhz official support. In addition, there are straps built into the motherboard that make overclocking either easier or harder. If I run 8*450 my FSB would be 1800Mhz. I realize that a faster FSB accomodates more data bandwidth. My question is how much performance increase do we get. My MSI board achieved a 504Mhz FSB on PCSTATS.COM
i have ran my balistix at both 1066 and 1200 and although there was an increase in bandwith shown in memtest around 250 mb/s the gains under windows use was minimal. my memory though needed 2.22 V to run at 1200 and 2.06 for 1066 , so i chose to run it at 1066 for the sake of my modules. strap used at 1200 was (266) and 1066 (400) - 9X400 with a quad
Your FSB is only 900MHz when runnig 8*450 (DDR2)
The most performance gain is from pure CPU overclock. You will see some gains when overclocking the ram to higher speeds using the dividers but not as significant as compared to the CPU OC. I prefer 1:1 ratio CPU>RAM ratio. Extreme OC'ers like the larger bandwidth of OC'd memory (using dividers/multipliers).
I get it now as far as FSB is concerned. I got confused while looking at CPU-Z with its bus speed and rated FSB. Essentially, my motherboard will support memory running at 1333Mhz. I don't think there's many DDR2 memory that can crank it up to 1333Mhz.Quote:
Originally Posted by sofarfrome
I ran 3DMark06 with 8*450 instead of 9*400. I saw no gains in my 3DMark score. Does cranking up the FSB increase the bandwidth of my system? Should I try for 7*500+ on the FSB? Will lowering my multiplier lower the CPU's temperature?
Well
I posted here my results at x7, x8, x9 multi...the lowest multi the more badwith in memories of course being concius of those fsb holes...
My quad has great performance at 450x8 than 400x9...
If i use everest or other tests, everything's faster at lower multis same speed.
Regards
Sergio
In synthetic tests and benchmarks as well as games and apps. you will certainly notice performance gains when OCing your memory. But if you run 7*500 (3.5GHz CPU x DDR2-1000 memory) vs. say 8*500 (4.0GHz CPU x DDR2-1000) you will see a more significant performance improvement than if you used dividers to achieve higher memory OC while keeping the same CPU OC. The idea is to optimize both memory and CPU OCs but keep in mind that the CPU should be a more important focal point as it will give you a better "bang for your buck".
I made a new record for myself with a Quadcore Q6600 and the Asus P5W DH Deluxe.
Its running @ 3,5Ghz 9x389 Multiplier Vcore 1,48 in bios, and 1,42 in windows
This is on air with a Arctic Freezer 7 Pro (not Lapped) Room temps around 22 a 24c
Have only test my memory to get it lower than 5.5.5.15
Becource it can be lower.
Here a screenshot of mine.
I will post later a stressed (stable testing) screenshot with the idle and stressed temps and the bios settings i use
http://members.home.nl/knight-rider-...203DMark05.JPG
Read this o/c guide re optimium settings and their effect. Very interesting and detailed. Download the speadsheet to try out settings and see what they give you. The impact of FSB , Bandwidth and cas latency will all become clear -lol!Quote:
Originally Posted by Hans Gruber
http://www.thetechrepository.com/showthread.php?t=195