Question about AMD CPU with unlocked cores

[Drakula]

Got a question about a Phenom II X2 550 BE that was unlocked to a Phenom II X4 B50. It is running at 3.1GHz, no overclock done. Left everything to either stock or auto in the BIOS setting. Had run OCCT and Prime95, both were ran for about 24 hours to ensure stability and both seems to end without problem. Want to use it for video encoding and most likely archiving some files, eventually might use it for some gaming. So the question is, would it be likely to produce some artifacts or corruption when encoding video due to unlocked cores?

Thanks in advance.

 

[dac7nco]

The 24hr Prime95 run cleared you of all ills... congrats and encode away!

Daimon

 

[tweakboy]

Yes do your thing. Your fine.

 

[frostedflakes]

The 24hr Prime95 run cleared you of all ills... congrats and encode away!

Daimon

Yeah give it a good beating with Prime95, 24-48 hours, something like that. If it gets through without any errors, I'd feel pretty confident about its stability.

 

[Drakula]

Thanks for the assurance everybody, it is a relief to hear that the system should be stable enough after running those tests. Barely do overclocking at all. For the previous systems, only run Memtest86+ and couple hours of OCCT to make sure the system is stable enough because everything is left to stock setting, this is pretty much the first time in a while this close to "overclock" the CPU, or so to speak. Last experience of overclock was the pencil trick of Athlon days. Anyway, will be preparing couple things to encode and save couple GB off the hard drive.

Thanks again!

 

[eternalone]

OCCT stresses the Cpu better, and destroys prime 95 so called stable systems. Give her an hour of OCCT and see if she is as stable as you say.

 

[frostedflakes]

On Intel systems maybe, but on AMD systems Prime95 seems to stress the processor as well as linpack or anything else out there. It's certainly not a bad idea to try multiple stress test programs to see if one picks up errors that others don't, though.

 

[nanaki333]

Yeah give it a good beating with Prime95, 24-48 hours, something like that. If it gets through without any errors, I'd feel pretty confident about its stability.

he said he already ran a 24 hour prime test

 

[xd_1771]

What motherboard is this running on? It must be safely capable of 125W processor operation

 

[Drakula]

OCCT stresses the Cpu better, and destroys prime 95 so called stable systems. Give her an hour of OCCT and see if she is as stable as you say.

Actually not sure which is better, but it was quite a while since putting together a new system. Did have OCCT problems back at the 2.x versions, it would always give some unknown error and abort the tests, but that changed with 3.x versions. Read about Prim95, as well as people mention it as a good stressing software, so thought gave it a try. OCCT did not killed the MSI 785GM-p45, but Prime95 did, so that was something.

On Intel systems maybe, but on AMD systems Prime95 seems to stress the processor as well as linpack or anything else out there. It's certainly not a bad idea to try multiple stress test programs to see if one picks up errors that others don't, though.

Yep, running multiple stress testing programs would not hurt. It might seem like waste of time and electricity, but as long as the system is stable, it is all worth it.

he said he already ran a 24 hour prime test

I am running another round test to further ensuring the stability due to change of Windows.

What motherboard is this running on? It must be safely capable of 125W processor operation

It is running on Gigabyte 790XTA-UD4, it is able to handle 125W ones without problem. They did state that it is able to handle 140W CPUs. The motherboard seems to be fairly sturdy and well equipped. Got it as a bundle from Fry's like a year or so ago when they were still offering them.


Have new questions about Prime95 tests, I am currently running another round of test, but unsure if using the correct test. Chose the "Blend" test because wanted to test out RAM as well. What are Small FFTs and In-place large FFTs tests best suited for? They both seem to stress CPU at some different level. From the description, it seems that larget FFTs test should be good for burn in, is that right? Please correct me if wrong.

Thanks.

 

[Drakula]

Sorry to bring this thread back. Previously, tried to unlock the same CPU with MSI 785GM-P45 but the motherboard when suicidal when successfully unlock the CPU with four cores and running Prime95. Now, after got the motherboard back from RMA, will attempt to put a CPU of 65W in it, is it fine to run Prime95? Or would the motherboard burn itself out again?

 

[Topweasel]

Sorry to bring this thread back. Previously, tried to unlock the same CPU with MSI 785GM-P45 but the motherboard when suicidal when successfully unlock the CPU with four cores and running Prime95. Now, after got the motherboard back from RMA, will attempt to put a CPU of 65W in it, is it fine to run Prime95? Or would the motherboard burn itself out again?

Prime can not kill a Mobo. Period end of story. What it can do is bring the systems components up closer (but no where near) to the TDP usage they are specced at. What isn't clear is what you were running in the MSI during the time it died.

If it was a X2 turned into an X4 and OC'd then its clear what killed it.

So yes a untouched 64w CPU should run fine in that board and yes you should be able to run Prime.

 

[Drakula]

Prime can not kill a Mobo. Period end of story. What it can do is bring the systems components up closer (but no where near) to the TDP usage they are specced at. What isn't clear is what you were running in the MSI during the time it died.

If it was a X2 turned into an X4 and OC'd then its clear what killed it.

So yes a untouched 64w CPU should run fine in that board and yes you should be able to run Prime.

At that time, I only unlock the additional cores. At first, it was not passing Prime95 or OCCT, so I crank up the core voltage slightly, as I have read someone's post on this forum that it makes the CPU stable.

 

[Topweasel]

At that time, I only unlock the additional cores. At first, it was not passing Prime95 or OCCT, so I crank up the core voltage slightly, as I have read someone's post on this forum that it makes the CPU stable.

It does. But you have to remember that the board had an issue with anything over 95w, you made it a 4 core processor with with it over volted. Trust me that CPU was closer to 125W then it was 65W.

Take it back. That board doesn't support a CPU putting out more then 95w. Yeah. I wouldn't even unlock cores on that. Thought it might be one of the touchy randomly blows up if you use a thuban board. But know that's one of those I am only made for Athlon II's or I explode boards.