New Here - Thoughts on my initial OC?

[RGone]

Maybe this is the reason why some of the workers are slow. Using myself as an example, when I'm trying to get X mhz oc stable ill set the Cpu V run prime, it fails 1 worker, I bump the Cpu V, it runs 2 hours and passes. Maybe it is because it is "just enough" voltage to keep the workers happy and working for 2 hours. I'm going to do an experiment when I have time and see if upping the Cpu V at a known stable oc will increase and/or have the amount of tests completed more inline with one another.

"Drake" I think you are on to something there. When you said low Vcore, it reminded me of that post I put up in one of those long test threads of mine. I benched Cinebench at one Vcore and then at a higher Vcore and the bench result went greater when I increased the Vcore. Such result lead me to believe at one point I was just barely using enough Vcore to run the bench at all.
RGone...

 

[SouthernStyle]

Maybe this is the reason why some of the workers are slow. Using myself as an example, when I'm trying to get X mhz oc stable ill set the Cpu V run prime, it fails 1 worker, I bump the Cpu V, it runs 2 hours and passes. Maybe it is because it is "just enough" voltage to keep the workers happy and working for 2 hours. I'm going to do an experiment when I have time and see if upping the Cpu V at a known stable oc will increase and/or have the amount of tests completed more inline with one another.

I'm doing this exact testing tonight myself. And trying to figure out why small Multi/HTT changes at the same VCore could cause different modules to swap places in performance. I'm curious to see what voltage if any will keep all workers balanced during testing.

 

[RGone]

I'm doing this exact testing tonight myself. And trying to figure out why small Multi/HTT changes at the same VCore could cause different modules to swap places in performance. I'm curious to see what voltage if any will keep all workers balanced during testing.

Just remember they are never all together. So a couple of tests ahead or behind has been normal for years. Not 10 vs 23 tests or so though.
RGone...

 

[SouthernStyle]

Just remember they are never all together. So a couple of tests ahead or behind has been normal for years. Not 10 vs 23 tests or so though.
RGone...

Right, I just want to see if I can achieve a respectable balance within reason. I just don't have much faith that my chip is of the special variety.

 

[RGone]

Okay I tested number of tests completed in the same amount of time and using VERY adequete Vcore.

Link here to the results >> http://www.overclockers.com/forums/showpost.php?p=7440401&postcount=148

I know MY results are not showing the workers to be too different in speed of the number of tests that are completed in the same amount of time. I know there was 'plenty' of Vcore. I know heat has us just barely using enough Vcore and that may result in the disparity in the number of tests completed on each core.

I know "mandrake4565" said he was going to test again on his tested stable speeds BUT add some Vcore and see if doing so brings his completing workers closer together.

Anyone with such disparity in completion of workers certainly needs to test for themselves. If you think you are P95 Blend stable at X amount of Vcore but workers are not completing fairly equally in the number of tests completing...then keep the same Vcore but lower the clock some and run P95 Blend mode. That would equal upping the Vcore (without adding heat) and then test to see if the workers all finish testing with the same general number of tests for each core.
RGone...

 

[Johan45]

That tends to make sense. The workers are struggling because they're starved for power.

 

[RGone]

That tends to make sense. The workers are struggling because they're starved for power.

Maybe and maybe not. Drake gave me the idea when he mentiioned starving for voltage. I was never more amazed than when I saw increasing my Vcore raised my CPU score slightly in Cinebench R11.5 CPU bench test.

I meant to use enough Vcore that voltage certainly would not be an issue. Now when I have time I will back down hoping to find that exact spot that completes P95 but shows the workers finishing far apart in the number of tests. Might see this and might not. Just a theory right now.

The more testers the merrier the chase.
RGone...

 

[Johan45]

I'll probably (with luck) be in my own little world most of the weekend. Got another identical 580 coming hope today or tomorrow. Got it thru the classies here from sobe.
So you see with a new toy gotta play!!!

 

[Mandrake4565]

Maybe and maybe not. Drake gave me the idea when he mentiioned starving for voltage. I was never more amazed than when I saw increasing my Vcore raised my CPU score slightly in Cinebench R11.5 CPU bench test.

I meant to use enough Vcore that voltage certainly would not be an issue. Now when I have time I will back down hoping to find that exact spot that completes P95 but shows the workers finishing far apart in the number of tests. Might see this and might not. Just a theory right now.

The more testers the merrier the chase.
RGone...

You can throw my theory out, I ran Prime with my 4.7 OC I lowered the Cpu V core 1 notch to where it was unstable previously. It failed prime after about 1 hour 51m to be exact but it did 34 tests on 3 cores and 35 on 5 cores.

 

[Johan45]

So they may just be weaker cores. That was kinda what I always though since for my experiece it was consistently the same ones that would fall behind. Didn't usually matter what frequency once I was over stock.

 

[SouthernStyle]

I'm still planning to do my testing. I said I was going to do it a few days ago but I've ended up having to use my PC for work....after I get home from work, and haven't had the chance to do it yet. And I've want to reseat my cooler before I start. But I'll be sure to post my results back here as soon as I can.

 

[SouthernStyle]

So they may just be weaker cores. That was kinda what I always though since for my experiece it was consistently the same ones that would fall behind. Didn't usually matter what frequency once I was over stock.

I always thought the same, but I have had cores trade places on being the weakest while testing for a stable OC. Like large, drastic changes in which core lags behind the most. But most of the time it is the same core (6). The strangest thing I notice is that whatever core is the slowest, the other core in that module will also be the fastest.

 

[RGone]

My core one one and five were the slightly slower ones but the rest swapped places seemed like all the time as to which would be faster.

Thanks for checking is out Sir Drake.
RGone...