Return to Level1Techs.com

Did I win the sillicon lottery?


#21

There is no heavier workload for a CPU I can think of than an hour of prime95 with small fft. If that passes with no failed workers then it should be stable in almost anything.


#22

Prime95? Sorry, never heard of it

Where can I find this?


#23

Found it, downloading as we speak.


#24

One question, with “failed” workers, do you mean “not running”?
I’m running Prime95 now, small fft, the temp of the CPU is up to 93C, all the cores are running but some of them change from “self-test” status to “not running”. This changes not often but some cores change sometimes.
Does this mean the overclock is not stable?


#25

As long as they continue to run then its fine. You will see worker failed and then the core will go idle and stay idle if the overclock is not stable.


#26

Until now no failed workers, CPU keeps at 93C, then it drops to 92C and back to 93C.
But all workers keep on “self-test” status, few workers (3) change sometimes to “not running”.
So, if this keeps un running for one hour with no failed/idle workers the overclock is stable?

I will keep it running for an hour then.


#27

check task manager to make sure all cores are 100%


#28

Test ran for almost 41 minutes and suddenly a blue screen.
So the overclock is not stable.
I will set it back to 5 Ghz


#29

I should tell you that prime95 is not a realistic workload. You could fail p95 tests but not encounter issues in day to day tasks. You could also pass p95 tests and fail doing other things.

Prime95 is good at one thing. Max core utilization and max heat production from a system.

I would say bump it to 1.365 and try again if you want, or just go back to 5ghz and try again.


#30

The CPU was running all the time at 94C, that is a lot for a 41 minutes I guess.
It was running ok, all the workers and no fails and suddenly a blue screen.
I will try it at 1.365 volts, if that is not working then I go back to 5Ghz.


#31

You got a good bin… usually these binned chips tend to cluster near the center of the wafer due to less impurities in the silicon being located there… HAHA nice grab man


#32

Hi,

Here is an update.

I changed to core voltage to 1.360 volts.
Ran Prima65 again, the temps got one degree higher then before, I ran the test for almost 55 minutes then I stopped it myself.
I don’t think that it is no longer needed, the test ran without failure of any worker, some workers changed status from “self-test” to “not running” and back.
For me this is a stable system.
I don’t think that my little brother will benifit from this 5.1 Ghz at 1.360 volts, I’m going to change it back to the 5Ghz and stable 1.35 volts and the system is stable enough for my brother, including temps.

many thanks again for the help and have a great new year.