• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

dumb question... could a CPU be partly damaged?

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

shiyan

Member
Joined
Aug 23, 2003
Location
behind the wheel
My Venice 3500+ seems to be acting weird so I was wondering... is it possible to only partly damage a CPU so that it scores less in benchmarks :eek:

At 1.55v, 2.9Ghz, it was giving around 12,200 MIPS in Sandra's Arithmetic bench.

Just tested it right now at the exact same settings, and it's not even 11k :confused:

With most programs closed CPU usage is at around 4% so it doesn't seem too strange :shrug:

I'm going to be making a fresh install in an hour or so, but was wondering if I might have "injured" the CPU by running P95 for around 6 hours at priority 10 (was using 1.5v and 2.8Ghz when I ran P95) :-/

Thermal throttling shouldn't be an issue as I'm using a XP90C with a Zalman 92mm which spins up to 2700rpm when temps get higher.

Please don't point and laugh, I feel dumb enough as it is :eek:
 
I believe that if it was damaged, it would just be unstable. Maybe you hit a setting in the bios you shouldnt have? or something like that? I wouldnt think you have damaged the cpu somehow just by running P95, unless it got too hot and i dont tihnk you're running enough voltage to get electromigration.

you can always send me that cpu and get a new one tho ;-) :D ill even send back my clawhamer 2800 in return. although its only s754..... :-/
 
Si sandra isn't great benchmark over some time, the score will be different each time and is effected by memory timings, some bios setting and what you have on a hard drive. If you reinstall windows your score should return to 12k, I guess but over time might decrease. If your cpu can prime then it's not damaged
 
dumb question... could a CPU be partly damaged?
I "partly damaged" a 1600 Pali once while cutting traces... Only had a multiplier of 6 after that, so anything could be possible i suppose :)
 
well, I managed to find the real problem - I hadn't disabled thermal throttling which was set at 50% in the BIOS and as a result when my room temperature was a little hotter than before the chip started throttling back in the middle of the Sandra test :p

man I feel dumb.... :eh?:

at least I've found the source of some weird issues I've had recently :)
 
Last edited:
how strange... now I'm not so sure it was due to thermal throttling any more... might be the RAID 5 array which is causing this

edit: yep, definitely the Sil 3114 RAID 5 causing the lower benchmarks - I'm guessing it was taking up CPU cycles trying to update the array :shrug:
 
Last edited:
Back