• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Prime95 Error

E

EhJayKim

Limp Gawd
Joined
Nov 18, 2004
Messages
361
So I keep getting the same error and I am wondering if anybody knows what could be causing it...memtest ran fine...So I don't think its memory, and its a rounding calculation...I have a venice 3200 on an xp-90 and at 2.55 ghz and above, I alaways get this error:

"FATAL ERROR: Rounding was 0.5, expected less than 0.4"

So, is it my CPU then? At, 2.55 it happened after 3.5 hours, at 260fsb it took 2 minutes. Thanks, oh, and any ideas on how to fix this? I can still get cpu-z to read my speeds too...
 
yah, it's the cpu...you probably need the bump the vcore up a bit on the cpu....otherwise keep it @ a safe speed hehe
 
It means your CPU is running at the speed...but not calculating correctly. You need to either add voltage (if your CPU is cool enough) or lower the speed until you get at least 24 hours stable....at least that's my opinion.
 
upped the vcore to 1.6 now, says temps are 41 at load...superpi stops when I try to do the 32M calculation...I'm going to let the whole system cool down to room temp, then try again...2.55...wish I could do better. 550Mhz isn't bad...I was just hoping for more :(. Getting got a windows error too...I think I can maybe add a tiny bit more vcore..but probably not too much...any other ideas?
 
yeah, it's definitly a cpu limitation.. either back off a bit, or aply more voltage. seeing as you're already at 1.6v, i would do the former ;)

also, i don't see your name in here yet :p
http://www.hardforum.com/showthread.php?t=888001
when you've finally got the stable limit of that cpu, go post with OPN on your heatspreader and the max speed/voltage ;)
 
darn... I was hoping for another thing to do...I really wish I could get a higher oc... any ideas? I mean, have I really hit the roof of the chip?
 
I've always been a fan of completely isolating the CPU when going for a higher overclock. You should take down RAM speeds and the HTT with dividers/multis so that the CPU is the only thing running higher then specs, then see if you do need more voltage. Also 24hrs. stable in prime95 isn't a necessity. I ran my system on 2min prime stable and no games or 3Dmark ever crashed, but its up to you.
 
EhJayKim said:
darn... I was hoping for another thing to do...I really wish I could get a higher oc... any ideas? I mean, have I really hit the roof of the chip?
Click to expand...

I wouoldn't say it is the limit of ur chip.....just the limited voltage/cooling it would get to get higher..I mean..I have an old ass athlon 64 3200+ clawhammer witht eh CO stepping...I can hit 2.23 ghz fully stable with 1.6 voltage..woohoo a 223 mhz overclock :rolleyes:
but If I bump the voltage up to 1.7...ouch..I can get 2.34 outta of the chip..but it is too hot hehe..so yah..get better cooling and bump the voltage up a bit..or be happy with what u have right now..2.5 isn't shabby hhehe
 
Well, could it be my ram if I dropped my cpu multi down to 8, then put my fsb up to the max my board supports (300mhz)? Also, I still can't find the HT multi, I have the ASUS A8V deluxe...giving the computer a chance to rest for a bit...see if it works... I had a problem...everytime I pushe dthe enter key in the bios, the whole thing froze, if it doesn't work when I try in the morning...i'll flash the bios. Also, I tried lowering the CPU multi, with it at 9, I got a little bit further, but with it at 9.5, I was able to get up to 2.66 Ghz, and still run SuperPI and CPU-Z. (Multi at 9 was something like 2.4Ghz). I had the fsb at 290 before the BSOD. I also had my voltage much lower...and it was running faster, so I think that it likes the 9.5 best. It has problems opening something called asmihwlo.dll on startup sometimes, it says its a kernel error, anybody know what this is? Also, any more suggestions? Thanks, Oh, and performance is best when I run with no RAM divider right? Even if the CPU is a little big slower...
Thanks
-A
 
Just read a thread about 1.6V killing a 3200+ so if i were you i'd be pretty careful. Im runnin mine at 2.5Ghz w/ 1.48V at the mo with load temps of 35ish C but it fails prime after 2 hours. Gonna bump to 1.5V when i get in from work to get a little more stability. Mem is runnin at 208Mhz with a 5:6 multi. Mem is at 2.7V hrs of memtest say thats cool. Cpu jus needs a little more juice i reckon ;)
 
Mr. K6 said:
I've always been a fan of completely isolating the CPU when going for a higher overclock. You should take down RAM speeds and the HTT with dividers/multis so that the CPU is the only thing running higher then specs, then see if you do need more voltage. Also 24hrs. stable in prime95 isn't a necessity. I ran my system on 2min prime stable and no games or 3Dmark ever crashed, but its up to you.
Click to expand...
well then, i hope you don't fold

while such a cavalier attitude to stability is fine for someone whose CPU calculations are not important, i.e. if you just game/watch videos/listen to music/browse web etc...

if you were doing important work (i.e. a DC project), then that would mean a lot -- if your CPU isn't doing math correctly, it's screwing up the DC project.

 
You must log in or register to reply here.
Back
Top