Same 930 chip, Gflops different on stability compared to other 930's.

[theveteran]

Hi all

I have been stability testing alot and have noticed that the same 930 chips give a different Gflops reading in Intel burn test and Linx,sometimes it 35 or 45 or 55 or 65...why is this?Are some of the chips faulty?
Thanks

[RVWinkle]

It's a stability test, not a benchmark.

[Anth Seebel]

gflops will indicate if your loading the cpu properly (stable settings).. if gflops are low the cpu is not actually loading properly at that clock , that is why gflops are low..
i have tested this extensively..
i am able to get higher clocks but gflops can be low which indicates to me that cpu is not properly loading - so il stop linx and adjust settings.. i can observe this with temps (along with low gflops).. a properly loading cpu will have a higher temp..

[theveteran]

thankyou for the explanation

[theveteran]

gflops will indicate if your loading the cpu properly (stable settings).. if gflops are low the cpu is not actually loading properly at that clock , that is why gflops are low..
i have tested this extensively..
i am able to get higher clocks but gflops can be low which indicates to me that cpu is not properly loading - so il stop linx and adjust settings.. i can observe this with temps (along with low gflops).. a properly loading cpu will have a higher temp..

How do i make the gpu load at he proper settings?
My theory is you cant unless you downclock?

[CryptiK]

Sometimes just stopping and restarting the test works for me.

[OC Nub]

Sometimes just stopping and restarting the test works for me.

Same for me, LinX is not a perfect program.

I can pass 25 loops LinX and still fail the latest version of p95 running blend in less than 1hr. IMO LinX is not the end all to stability testing.