Jump to content

Opteron oc not stable


Recommended Posts

well i got my replacement g skill sticks yesterday after about 3 rma's :( , i also got my new opteron 170 0550 UPMW stepping.

 

i have been trying to overclock it today, but getting prime 95 to give out rounding errors again.

 

3 cpu's i have gone through now and still errors.. this is so annoying, can anyone help?

 

i used the proper settings for the G skill 4400 LF's at 275mhz frequency, also 510B bios that i flashed to after installing my chip, 7-10x multiplier tried and usually prime95 has failed within a few minutes.

 

So what can i do now? am i stuck and cannot oc for the 3rd time, if i have bad hardware i cant believe it.. cos ive gone through quite alot.. this is bs :(

Share this post


Link to post
Share on other sites

  • Replies 39
  • Created
  • Last Reply

Top Posters In This Topic

So you are at the max cpu multi and the thing wont overclock correct?

 

 

 

What are your temps on that stock heat sink? What TIM are you using and waht does it look like when you take it off if you have.

 

 

Most inportant what volts are you using not everyone gets lucky and is able to overclock at stock volts.

Share this post


Link to post
Share on other sites

Why not use the 704-2BTA Bios.

Have you warmed the opty up yet.

I don't think the 510 is setup for Dual core.

Take it easy, it takes patience to OC an Opty, well, usually.

Perhaps I misunderstand you, but it sounds like, 'I just got my new dual core opty and new GSkill Ram and they aren't stable 2.75 GHz'.

 

Are you perhaps expecting too much?

Share this post


Link to post
Share on other sites

Yes, you need to update your bios to at least the 623-3 version to support dual core. This is most likely the root cause of your problem. That stepping of opty (I should be getting the same today) is golden and people are hitting 3ghz+ on it.

Share this post


Link to post
Share on other sites

i have been trying to overclock it today, but getting prime 95 to give out rounding errors again.

 

3 cpu's i have gone through now and still errors.. this is so annoying, can anyone help?

 

i used the proper settings for the G skill 4400 LF's at 275mhz frequency, also 510B bios that i flashed to after installing my chip, 7-10x multiplier tried and usually prime95 has failed within a few minutes.

 

So what can i do now? am i stuck and cannot oc for the 3rd time, if i have bad hardware i cant believe it.. cos ive gone through quite alot.. this is bs :(

I've been getting rounding errors on my opteron 165 as well (only on cpu0 core though, some others reported similiar issues). It could be that some opterons are more sensitive to htt/mult/timings/etc than others. It could also be a prime95 issue (yeah yeah flame away. I'm a software dev. Things don't always work on every piece of hardware).

 

Before I left for work, I loaded optimized defaults, increased vcore a bit and set two primes to run. That should tell me if it is due to oc settings.

Share this post


Link to post
Share on other sites

no 1 has the slightest idea? my chipset usually is about 50c when stressing btw, would this cause errors?

 

When Prime craps out like that there are only 3 reasons...

 

1. The memory has gone beyond its limitations

 

or

 

2. The CPU has gone beyond its limitations

 

or

 

3. The htt (FSB) has gone beyond its limitations

 

Now, what you have to do is isolate what the bottleneck is. Start with the FSB. Tweak the multi and divider for the CPU and mem so that they run AT or BELOW stock settings for whatever FSB you are going to use. Move the FSB up slowly and stress it with dual primes for about 3-4 hrs on each increment so you can see where it needs more voltage. Give it juice, crank it up, prime it again. Do this until you go as high as you need to stable (or as high as you are comfortable with the voltage being). This will give you a good idea of the limitation of your htt.

 

Next, do the memory. Repeat this procedure using memtest (10 passes of test 5 and test 8 should be sufficient to determine stability). Run the memory faster and faster until you start getting errors. Once you get no more errors at your highest speed, then go into Windows and Prime them some more. You may need to back them off a bit from memtest stability to make them Prime stable. Then you know the limitation of your memory.

Tip: I have the G. Skill PC4000 3-4-4-8 matched pair. If yours are anything like mine (and they probably are), then if you can't get a speed you want on 2.6V, then you're not going to get it by raising the voltage. Raising the voltage on mine actually makes them crap out faster and error worse.

 

After you find these 2, then you can overclock your CPU and know exactly how much juice to give your FSB for whatever speed you want to get it to, and know exactly where you will need to use a divider for your memory so you don't corrupt your data. Then just Prime the heck out of the CPU to ensure stability. If prime craps out then up the voltage.

 

This process can sometimes take between 2 weeks to 2 months depending on how much time you are able to devote to it on a regular basis. Just be patient and be very thorough. That is what makes a good stable overclock. Good luck man!! :cool:

Share this post


Link to post
Share on other sites

ok, i did what you said.

 

I put memory speed 270mhz all the correct timings given by gskill, put dram strength to 5 and the other strength to 2. Voltage to 2.8 & proc speed only oced by about 60mhz over stock, i got 73 passes of test 8 and 2 failures.

 

PMW temp was 57c after looking in bios on reboot

Share this post


Link to post
Share on other sites

ok, i did what you said.

 

I put memory speed 270mhz all the correct timings given by gskill, put dram strength to 5 and the other strength to 2. Voltage to 2.8 & proc speed only oced by about 60mhz over stock, i got 73 passes of test 8 and 2 failures.

 

PMW temp was 57c after looking in bios on reboot

 

Some of your timings probably aren't quite right... Loosen some of the timings not specified by the SPD. I know on my G. Skills the stability is affected a lot by the trc and trfc (mine are 16 and 12 respectively). You should be able to tweak the non-SPD timings to make them run @ SPD timings and 275MHz memtest stable. Also, try keeping the voltage @ 2.6V because the G. Skill don't like the extra voltage.

Share this post


Link to post
Share on other sites

196 passes and 8 errors of test 8, ran for 7 hours 49 minutes.. @ 275mhz

 

PWM temp was 57C also chipset was around the same i think..

 

oh and i had to put voltage of ram up to 2.7 otherwise it wouldnt boot that far.

 

then i had errors early on that voltage so i went to 2.8..

 

what now? nearly 8 hours of just test # 8 and it fails :confused:

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...