Prime95 fails after 12.5 hours? Should I up the voltage again or is it stable?

Hey,

Just a quick question. OC'ed processor from 2500 to 3200. Upped the
voltage to 1.7 because of Prime95 failing after a few minutes. At 1.7 volts,
Prime95 ran for 12.5 hours and then stopped with a rounding error.

Should I up the voltage again, or if it ran for that long, does it mean it's
stable, and that was some sort of fluke?

I'm pushing the temperatures though, so I'm a little hesitant. It's as low
as 45 degrees Celsius when idle, but as high as 57 degrees under full load.

What would you guys do? Should I instead lower the FSB by 10 or so?

Thanks for the advice!

hi

i have a 2500 @ 2300 and i got some errors also in prime95 but after i
updated to last version they dissappeared. (23.7.1)
you may try to lower fulload temp. i would continue with the test before
upping vcore any more, if only few errors show up some hours apart
then.....depends on what you want! errorfree or just stable enough.

coco

"Eternally" skrev i meddelandet
...
Hey,

Just a quick question. OC'ed processor from 2500 to 3200. Upped the
voltage to 1.7 because of Prime95 failing after a few minutes. At 1.7
volts,
Prime95 ran for 12.5 hours and then stopped with a rounding error.

Should I up the voltage again, or if it ran for that long, does it mean
it's
stable, and that was some sort of fluke?

I'm pushing the temperatures though, so I'm a little hesitant. It's as
low
as 45 degrees Celsius when idle, but as high as 57 degrees under full
load.

What would you guys do? Should I instead lower the FSB by 10 or so?

Thanks for the advice!

"CoCo" wrote in message
...
hi

i have a 2500 @ 2300 and i got some errors also in prime95 but after i
updated to last version they dissappeared. (23.7.1)
you may try to lower fulload temp. i would continue with the test before
upping vcore any more, if only few errors show up some hours apart
then.....depends on what you want! errorfree or just stable enough.

coco


Well I guess it depends on what this error really means. Does it mean that
if I run a program which uses the CPU full load for longer then 12 hours,
it'll probably crash? If so, then it's not stable enough. I want to be
able to encode video and stuff which can take that long.

If it doesn't mean that, then what does this prime95 rounding error really
translate into?

There are no 'real world' programs that would hold your CPU at full load for
12 hours...so you be pretty sure there wont be a problem

--
*****Replace 'NOSPAM' with 'btinternet' in the reply address*****
"Eternally" wrote in message
...
"CoCo" wrote in message
...
hi

i have a 2500 @ 2300 and i got some errors also in prime95 but after i
updated to last version they dissappeared. (23.7.1)
you may try to lower fulload temp. i would continue with the test before
upping vcore any more, if only few errors show up some hours apart
then.....depends on what you want! errorfree or just stable enough.

coco


Well I guess it depends on what this error really means. Does it mean
that
if I run a program which uses the CPU full load for longer then 12 hours,
it'll probably crash? If so, then it's not stable enough. I want to be
able to encode video and stuff which can take that long.

If it doesn't mean that, then what does this prime95 rounding error really
translate into?

hi

prime95 use continous floatingpoint calculations and checks the in and out
values vs the rest from the operation all the time and will detect nearly
any kind of computing error. this is a continous operation and video
encoding is not! (video encoding values does not depend on each other for
hour after hour)

coco

"Eternally" skrev i meddelandet
...
"CoCo" wrote in message
...
hi

i have a 2500 @ 2300 and i got some errors also in prime95 but after i
updated to last version they dissappeared. (23.7.1)
you may try to lower fulload temp. i would continue with the test before
upping vcore any more, if only few errors show up some hours apart
then.....depends on what you want! errorfree or just stable enough.

coco


Well I guess it depends on what this error really means. Does it mean
that
if I run a program which uses the CPU full load for longer then 12 hours,
it'll probably crash? If so, then it's not stable enough. I want to be
able to encode video and stuff which can take that long.

If it doesn't mean that, then what does this prime95 rounding error really
translate into?

Then why don't you try some video encoding over night and see what happens?

Dave


"Eternally" wrote in message
...
"CoCo" wrote in message
...
hi

i have a 2500 @ 2300 and i got some errors also in prime95 but after i
updated to last version they dissappeared. (23.7.1)
you may try to lower fulload temp. i would continue with the test before
upping vcore any more, if only few errors show up some hours apart
then.....depends on what you want! errorfree or just stable enough.

coco


Well I guess it depends on what this error really means. Does it mean
that
if I run a program which uses the CPU full load for longer then 12 hours,
it'll probably crash? If so, then it's not stable enough. I want to be
able to encode video and stuff which can take that long.

If it doesn't mean that, then what does this prime95 rounding error really
translate into?

On Wed, 04 Feb 2004 17:39:21 +0000, Eternally wrote:

Hey,

Just a quick question. OC'ed processor from 2500 to 3200. Upped the
voltage to 1.7 because of Prime95 failing after a few minutes. At 1.7 volts,
Prime95 ran for 12.5 hours and then stopped with a rounding error.

Should I up the voltage again, or if it ran for that long, does it mean it's
stable, and that was some sort of fluke?

I'm pushing the temperatures though, so I'm a little hesitant. It's as low
as 45 degrees Celsius when idle, but as high as 57 degrees under full load.

If 57C socket temp. that's too high when trying to reach higher speeds. If
it's the core temp, then it's probably ok.

What would you guys do? Should I instead lower the FSB by 10 or so?

I don't know. Do you have a UPS? Could a power glitch have caused the
problem? All my machines are behind a ups. If you value your data, you
will have one too. What cpu cooler do you have? If it's the stock cooler
it isn't good enough with the increased speed/voltage. At minimum, get
something comparable to a Vantec Areoflow. What's your case temp? Ambient
room temp? All these will affect cpu temp and the stability of the system.
400WPS? Are you sure your ram is stable at 200Mhz? Etc. 12hours without an
error is pretty good, but I'd want to know what caused the error even if
it was just one, and about all you can do is try and find the weak spot.
More voltage may fix it, but you need to keep temps down too. Good luck.


--
Abit KT7-Raid (KT133) Tbred B core CPU @2400MHz (24x100FSB)
http://mysite.verizon.net/res0exft/cpu.html

"Eternally" wrote
Just a quick question. OC'ed processor from 2500 to 3200. Upped the
voltage to 1.7 because of Prime95 failing after a few minutes. At 1.7
volts,
Prime95 ran for 12.5 hours and then stopped with a rounding error.
Should I up the voltage again, or if it ran for that long, does it mean
it's
stable, and that was some sort of fluke?
I'm pushing the temperatures though, so I'm a little hesitant. It's as
low
as 45 degrees Celsius when idle, but as high as 57 degrees under full
load.


Hi,

everyone has different standards, but I myself like to run Prime95 for a
minimum of 24Hours (48 if I'm away). I have a 2500+ @ 2.2GHz but it won't
last 24 hours *Priming* without 1.775vCore. My max load temp so far has
been 58°c.

One thing you can do is make sure you have a decent cooler and also use a
good thermal-paste. One things that helps allot is ensuring your case has
proper ventilation. If you can bring your case temps down by 5°c, you will
see that your CPU temps drops by 5°c also. Allot of people run their
overclocked rigs with the doors off (or completely spread out on a table!).

Here is the spec of my last AMD machine (just sold!)
--
Wayne ][

Barton (AQXEA) XP2500+ @ 2.2GHz (10x220) - 1.775vCore
CoolerMaster Aero 7 Lite - 3,200rpm
ABIT NF7-S v2.0 (BIOS d20)
512MB Dual TwiSTER PC3500 @ DDR440 1:1 (2.0,3,3,9 - 2.8v)
Sapphire Atlantis 9800 - 3.3ns Samsung (325/290 Default)
WD-SE 240GB (2x120GB) SATA RAID-0 (NTFS - 16k Stripe)
Antec SX630II Mini-Tower Case Inc 300w PSU
2 x CoolerMaster 80mm Blue Neon Fans
WinXP PRO inc. SP1
nVidia Unified v3.13
Cat 3.7 - DX9.0b

prime95 for 12 hours is great. my pc is rock stable (on 24/7 and i play a
lot of grafx intense games. prime95 fails on my always 45 min- 1 hr. so draw
ur own conclusions
"Eternally" wrote in message
...
Hey,

Just a quick question. OC'ed processor from 2500 to 3200. Upped the
voltage to 1.7 because of Prime95 failing after a few minutes. At 1.7
volts,
Prime95 ran for 12.5 hours and then stopped with a rounding error.

Should I up the voltage again, or if it ran for that long, does it mean
it's
stable, and that was some sort of fluke?

I'm pushing the temperatures though, so I'm a little hesitant. It's as
low
as 45 degrees Celsius when idle, but as high as 57 degrees under full
load.

What would you guys do? Should I instead lower the FSB by 10 or so?

Thanks for the advice!

"Eternally" wrote in message
...
| Hey,
|
| Just a quick question. OC'ed processor from 2500 to 3200. Upped the
| voltage to 1.7 because of Prime95 failing after a few minutes. At 1.7
volts,
| Prime95 ran for 12.5 hours and then stopped with a rounding error.
|
| Should I up the voltage again, or if it ran for that long, does it mean
it's
| stable, and that was some sort of fluke?
|
| I'm pushing the temperatures though, so I'm a little hesitant. It's as
low
| as 45 degrees Celsius when idle, but as high as 57 degrees under full
load.
|
| What would you guys do? Should I instead lower the FSB by 10 or so?
|
| Thanks for the advice!
|


12.5 hours isn't bad, but for a perfectly stable, properly configured
system, it should run the test infinitely with no problems. I would think,
though, that it is unlikely you would be taxing your CPU for that hard for
that long at any time, so yo will likely be ok. My personal satisfaction
level is 24 hours, since I do some rendering that keeps the CPU busy for
12-15 hours at times, but that's just my picky personal preference.