YA(Opteron 165)T

[arcenite]

Before burning in I was able to get this chip to 2.6ghz without upping the voltage. Right now I am burning in at stock speeds @ .992v. Should I expect any kind of OC increase or am I doing this for my health? Thanks

Bill

Update: Achieved 2.7ghz @ 1.392v. Should I try 2.8? Prime testing now...

Update2: Tested stable @ 2.7/1.392. Tried 2.75/1.392 and failed. Now trying 2.75/1.408... Passed, staying at 2.7ghz due to voltage and heat issues...

Update3: I decided to push it a little further... After having issues with S&M (says core1 fails at the FPU test even at default speed/voltage) I decided to ignore it because SuperPI and Prime95 have no issues as of yet. I just passed SuperPI @ 2.8ghz/[1.488(cpuz)][1.525(bios)]. I am priming to see what the deal is... I reached my goal of a 1ghz overclock though

 

[Wentelteefje]

Theoretically, it should overclock higher... In real life, go for it!

BTW which program do you use? CPU Burn In?

EDIT: Very impressive BTW!

 

[arcenite]

Originally posted by: Wentelteefje
Theoretically, it should overclock higher... In real life, go for it!

BTW which program do you use? CPU Burn In?

EDIT: Very impressive BTW!

I am using Prime95 x 2... and thanks

 

[Wentelteefje]

Cool... How long have you been burning in for this result?

 

[arcenite]

I started about 5 minutes ago

 

[Wentelteefje]

FTW... You're lying...

 

[RichUK]

To be honest burn in is just a myth, and it rarely works if ever. I personally just think it is a load of crap, since the manufacturers already test the chips more vigorously then any apps that we can download.

But if you want to get a toasty core just for the fun of it use S&M. This will turn any AMD64 into a Intel fire breathing PresHott straight off the bat.

 

[Wentelteefje]

My favourite is OCCT... AFAIK this is about the most stressing it can get... If burning in is a myth, how come the guy can let his Opty run @ 0.992V...?

 

[arcenite]

Originally posted by: Wentelteefje
My favourite is OCCT... AFAIK this is about the most stressing it can get... If burning in is a myth, how come the guy can let his Opty run @ 0.992V...?

Well the truth is, I just started burning in not long before this post... But I was running at 2.6ghz and quite often ran prime through the night to make sure there were no problems.. I suppose that can be considered burning in...

 

[Wentelteefje]

K... Thanks, I'll give it a try on my Sempron... I want to hit 2.6GHz with it...

 

[RichUK]

Originally posted by: Wentelteefje
My favourite is OCCT... AFAIK this is about the most stressing it can get... If burning in is a myth, how come the guy can let his Opty run @ 0.992V...?

Because processors have different yields, hence why some single core opterons can reach 3Ghz on near stock volts, where as some SD cores need 1.6v to reach 2.7Ghz. Although in some cases people claim it has worked, and when I say worked it hasn?t allowed them to OC higher but simply supply less voltage at its highest stable frequency. Well, so the claims go. How true they are I don?t know. I am just a none believer, and also this discussion has been brought up many times before, with no conclusive answer (aka no scientific answer), not just on this forum.

P.S it is not uncommon to run a chip at stock speeds undervolted, it just brings us back to the processor yield subject.

 

[Wentelteefje]

Originally posted by: RichUK

Originally posted by: Wentelteefje
My favourite is OCCT... AFAIK this is about the most stressing it can get... If burning in is a myth, how come the guy can let his Opty run @ 0.992V...?

Because processors have different yields, hence why some single core opterons can reach 3Ghz on near stock volts, where as some SD cores need 1.6v to reach 2.7Ghz. Although in some cases people claim it has worked, and when I say worked it hasn?t allowed them to OC higher but simply supply less voltage at its highest stable frequency. Well, so the claims go. How true they are I don?t know. I am just a none believer, and also this discussion has been brought up many times before, with no conclusive answer (aka no scientific answer), not just on this forum.

P.S it is not uncommon to run a chip at stock speeds undervolted, it just brings us back to the processor yield subject.

Your explanation seems right... I'll give it a try nonetheless... See what it gives... Thanks!

 

[arcenite]

Update

 

[Ronnie]

Thats a really nice 165, what stepping?

 

[nealh]

Originally posted by: RichUK

Originally posted by: Wentelteefje
My favourite is OCCT... AFAIK this is about the most stressing it can get... If burning in is a myth, how come the guy can let his Opty run @ 0.992V...?

Because processors have different yields, hence why some single core opterons can reach 3Ghz on near stock volts, where as some SD cores need 1.6v to reach 2.7Ghz. Although in some cases people claim it has worked, and when I say worked it hasn?t allowed them to OC higher but simply supply less voltage at its highest stable frequency. Well, so the claims go. How true they are I don?t know. I am just a none believer, and also this discussion has been brought up many times before, with no conclusive answer (aka no scientific answer), not just on this forum.

P.S it is not uncommon to run a chip at stock speeds undervolted, it just brings us back to the processor yield subject.

this has been confirmed by several of the intel engineers on AT ..ie that burn in is a myth

more likely just testing and tweaking, allowing the system to settle in

 

[arcenite]

Originally posted by: Ronnie
Thats a really nice 165, what stepping?

It is a CCBWE 0546. I am not sure of the rest of the stepping as I did not record it before I put it in. I was in too much of a hurry.

 

[Tu13erhead]

Does burn-in really help overclocking?

 

[Lord Banshee]

where did you buy it from?

 

[arcenite]

Originally posted by: Lord Banshee
where did you buy it from?

Monarch... Another update...

 

[RichUK]

Your results are good, very good.

But as a result of burn-in i doubt it.

 

[SophalotJack]

I have a question about prime testing the opteron 165. Do you all run two prime95's (one fo reach core) to max out the cores? Or is there a way to have one prime95 app max out the two cores?

 

[RichUK]

Originally posted by: SophalotJack
I have a question about prime testing the opteron 165. Do you all run two prime95's (one fo reach core) to max out the cores? Or is there a way to have one prime95 app max out the two cores?

Yes one instance per core, since prime95 is a single threaded app. Unlike S&M, which is multithreaded which i quite like to use now.

 

[arcenite]

Originally posted by: RichUK

Originally posted by: SophalotJack
I have a question about prime testing the opteron 165. Do you all run two prime95's (one fo reach core) to max out the cores? Or is there a way to have one prime95 app max out the two cores?

Yes one instance per core, since prime95 is a single threaded app. Unlike S&M, which is multithreaded which i quite like to use now.

What is this S&M? And Rich, what are your temps at that voltage? Thanks..

Edit: I found the S&M program and am going to use it per your recommendation... I am also keeping my frequency at 2.7Ghz cause anything higher requires more voltage...

Bill

 

[RichUK]

Originally posted by: arcenite

Originally posted by: RichUK

Originally posted by: SophalotJack
I have a question about prime testing the opteron 165. Do you all run two prime95's (one fo reach core) to max out the cores? Or is there a way to have one prime95 app max out the two cores?

Yes one instance per core, since prime95 is a single threaded app. Unlike S&M, which is multithreaded which i quite like to use now.

What is this S&M? And Rich, what are your temps at that voltage? Thanks..

Edit: I found the S&M program and am going to use it per your recommendation... I am also keeping my frequency at 2.7Ghz cause anything higher requires more voltage...

Bill

S&M is a multithreaded app, that allows you to test the stability of both cores with a single application. This app is very intensive and can be ran in either a loop or a certain amount of time. Also I perceive this app as good as testing with ?dual prime/stress prime 2004?, although I do use both to test stability, this app will also produce the most heat from your processor so be aware.

At 2.6 Ghz with 1.52v I load at around 52-53 degrees IIRC, this is with a xp90 with 47cfm 92mm panaflo fan. I can also reach 2.7Ghz on the same volts stably with around the same load temps.

 

[arcenite]

Originally posted by: RichUK

Originally posted by: arcenite

Originally posted by: RichUK

Originally posted by: SophalotJack
I have a question about prime testing the opteron 165. Do you all run two prime95's (one fo reach core) to max out the cores? Or is there a way to have one prime95 app max out the two cores?

Yes one instance per core, since prime95 is a single threaded app. Unlike S&M, which is multithreaded which i quite like to use now.

What is this S&M? And Rich, what are your temps at that voltage? Thanks..

Edit: I found the S&M program and am going to use it per your recommendation... I am also keeping my frequency at 2.7Ghz cause anything higher requires more voltage...

Bill

S&M is a multithreaded app, that allows you to test the stability of both cores with a single application. This app is very intensive and can be ran in either a loop or a certain amount of time. Also I perceive this app as good as testing with ?dual prime/stress prime 2004?, although I do use both to test stability, this app will also produce the most heat from your processor so be aware.

At 2.6 Ghz with 1.52v I load at around 52-53 degrees IIRC, this is with a xp90 with 47cfm 92mm panaflo fan. I can also reach 2.7Ghz on the same volts stably with around the same load temps.

Yes I noticed that about the program's ability to get things heated up. I will stick to this program from now on, thanks

Bill