I took IHS off mine after 3 bad mounts on IHS. It's so damn hard to mount with an IHS :(
Printable View
I took IHS off mine after 3 bad mounts on IHS. It's so damn hard to mount with an IHS :(
You guys may have chips with poor IHS contact. I'd suggest an RMA if you have unusially high temps at stock voltage.
Heres a screenie to go with that CPUz validation:
http://i26.photobucket.com/albums/c1...Wreck/3gPI.jpg
Again, is it dangerous to dual prime at 55c load?
For now, at least.
when windows reboots while loading does that mean i need a bigger PSU?
Currently have a 380 W antec PSU, and get reboots at 2.4 ghz on a opty 165 CCBWE 0534
Just wondering if its most likely mob (epox 9nda3j) or PSU
safeQuote:
Originally Posted by edma2
I just ran Tcasemax and my chip is 49C :stick: Bit low huh? I got a T on the substrate and an H inside the IHS... figured this thing would be good but nope, only 2.75 stable :(
definatly the psuQuote:
Originally Posted by socrilles
Quote:
Originally Posted by afireinside
yeah i got a 180 0530TPMW that has the maxTcase of 49c also but mine seems to over clock just fine. right now im priming at 12x260 3120mhz 1.5v in bios and reading 1.45v in windows.. im not so sure this max Tcase temp thing is what they claim it to be theres got to be more to it...
btw: ive got the T on my substrate also havent removed the ihs so have no idea whats under that...
http://www.ausfx.com/gallery/albums/...0in%20bios.JPG
peace perc,
i got my 170 @ 2.4ghz doesn't go over mid-30s Celsius when under load @ 1.42v (i just upped the voltage anyway cuz i was trying to oc higher).
now could someone point me to some place i could figure out how to use the memory dividers on this board :) i read all about the LDT/FSB stuff but nothing i use will work, except Auto. (i hit 2.4 by just putting FSB up to 240 and leaving dram at SPD)
i cant explain it any better. the ram speed of a64 systems is derived from the cpu speed because of the integrated mem controller. the ratios in the bios are just an approximation.Quote:
Originally Posted by Otaking71
2.75ghz 1.55V ran 38C solid dual D2OL overnight 100% stable :woot: I wish core 1 didn't suck so much but damn this is a nice chip! Now I need a new raptor... I broke it and I'm on a 7200 RPM 2mb cache HDD. Kind of killing the whole dual core experience :(
that helped a lot. i think my problem was the HTT range. whats the limits to be able to boot?
right now i'm at:
FSB: 320
HTT: 320x3 (640x3?)
RAM: 220ish (i dunno exactly, i put a 7/10 divider on)
CPU: 320x8.5=2720
vcore: 1.52 reading in smartguardian, i set 1.55 in bios
i have two instances of prime95 running for each core with the 'max heat and power consumption' test on each. temperature in smartguardian is beginning to flirt with 40 C after a few minutes in the high 30s. so far so good... (XP-90C with 92mm Vantec Tornado @ 4400rpm)
i'd rather run the FSB lower though :\ but the friggin' HTT won't let me boot cuz it winds up way out of range :X aren't there any better multipliers for HTT/LDT, like some kind of special control
btw i have a CCBWE 0534SPMW Opt 170
BAH.
tried 330x8.5 (2805) and core 0 fails prime95 within a minute :\ core 1 kept going indefinitely. (was at 1.56 vcore in smartguardian, 1.55x102.5% in BIOS). temps were still barely under 40 C. should i just give up or feed it more voltage?
From what I've heard x64 is not .5 friedly. Set your multiplier to a whole number.Quote:
Originally Posted by SubZeroGTS
i can't...if i use 9, then i have to use 300 FSB to hit 2700 or higher. and for LDT, 300x4x2=2400 HTT, and 300x3x2=1800 HTT, and both of those won't boot.
if i use 8, then i'll need a really high fsb... like 337-340...can the board handle that? and i'd still be hitting around 2.7ghz on this chip anyway (i haven't seen people with CCBWE 0534SPMW really get too far over that, i can boot into 2.8 range but just can't prime stable)
what sucks is i have to use a 7/10 divider on my ram so it ends up at 224MHz when its native 233MHz. so if anyone has experience with this board (sli-dr expert) at 330-350 FSB, and it can handle it...i might try that, just to get my ram speed up
333 htt * 3 is higher than 300 * 3 buddy... I don't know where you're getting the 2 from.
300 htt
3x ldt
9x multi
That's what you want. board will do 350+ htt.
300x3 LDT won't boot
i will experiment with higher fsbs then. i can't believe it can be so stable approaching 350 fsb? :o should i mess with chipset voltage at all for 340-350 fsb
(edit: tried 350x8 and core 1 failed after 10 min :\...smartguardian reading 1.6 vcore)
2nd edit: ok i'm now at 340x8 since 2720 seems to be the limit of my chip (1.53 vcore in smartguardian, im thinking i will need to bump that back up a little) i'm running my ram at 247MHz 2.5-4-3-7-1T @ 2.92v (the next divider gave me 225MHz, 2-3-2-5 but this ram is warranteed up to 2.9v)
2.9 GHz 32mb SuperPi
http://img506.imageshack.us/img506/9...erpi0ox.th.jpg
i'm doing some initial testing.....2.5GHz 1.4v prime so far....i'm not a prime person but i see everyone :banana::banana::banana::banana::banana:ing about it.........i think temps are out a bit....look at the load temps :slobber:
http://img513.imageshack.us/img513/1...z14v34m9aq.jpg
yeah i hear you man im not a prime person either but i do it to test for stable over clocks. hear is a shot of my optie 180 primming away at 3gigs seems stable so far...
btw: what bios are you guys using with these dfi ultra-d boards to get it to reconise these opties correctly? im sick of seeing optie 280 series on a 940 pin platform...
http://www.ausfx.com/gallery/albums/...gigs%201hr.JPG
peace perc,
Perc, can you run a sandra bandwidth quick and show me a screenie of your bandwidth efficiency?
yeah ill run it when i get a chance. right now im primming away and probly wouldnt get an accurate score. i ran it earlyer and it was 7100ish at 250htt and 89% eff but i havent got my tccd tuned in yet i was getting 91% with my fx57 and ballistix but that was a hole other system... ill post it for you later or ill pm you with the screenie..
peace perc,
great CPU perc
what sort of cooling are you on
i bumped the voltage a bit....2.6GHz ATM...i'm happy with this for 24/7 :toast:
http://img511.imageshack.us/img511/1...142v33m7qj.jpg
Perc is running a phasechanged rig... I'll have my 180 in on monday under a nice custom water setup, we'll see how she clocks with a more common form of cooling :)
My Opteron 170 @ 2.9 GHz can pass SuperPi 32mb easily but fails Prime no matter what the voltage is. (Fails Prime in about 2 Hours) My CPU load temps don't seem that bad, but I guess I have reached it's limit. (For my system anyway)
ok i am now priming at this:
340x8=2720 @ 1.55 vcore (stable 1.55 reading in both smartguardian and cpuz no fluctuation under load)
chipset voltage @ 1.53 in smart guardian
max load temp = 40 C (if the Vantec gets over 4000rpm it stays under 39 C)
(this is Opt 170, 0534)
that other guy said he got 2.8 dual prime stable @ 1.55v with this stepping, so i guess i'm close enough? one of my cores failed after an hour (320x8.5 @ 1.53vcore) so i'm at this voltage now. i also raised the FSB in order to use a 8.0 multiplier instead of 8.5 since i read whole number multipliers are more stable??? is this true?
i've noticed the SLI-DR Expert seems to have a narrower range of stable HTT. my system will NOT boot (or lockup if it does) if it is out of the 1910-2090 range.
(EDIT: agh, core 1 failed this time for a change after 32 minutes :\ i just backed it down to 338x8 (2704MHz) and this time core 1 failed within a minute. guess that bit about being stable at even multipliers was a myth, i'll go back to the 320x8.5)