Jump to content

X2 4400+: One core error at 15.5 hours! BAH!


keysersoze

Recommended Posts

So I kind of gave up on hitting 2500MHz with my X2 4400+ on air cooling, because my 2nd core kept erroring after a little over 2 hours while the first core would keep going for at least 8 hours.

 

I brought the FSB down to 240 X 10 and tried 2400MHz on 1.425 CPU VID. Both cores ran without error for 9hrs 15min. But CPU temp was up around 48 full load on a Thermalright XP-90 and Vantec Tornado (being controlled by my PSU, not on full blast), so I decided that I could probably get 2400MHz dual Prime stable at a lower voltage.

 

I brought the CPU VID down to 1.375 and my temps dropped to 45 degrees full load. The 2nd core went 24 hours with no errors....but the 1st core errored on me at 15hrs 30min!!!!!! Grrrr.....

 

Anyway, would you guys recommend that I bump the CPU VID to 1.400 and Prime again? Also, please look at my settings below (for the Prime that just failed on core 1 at 15.5 hours), and tell me if anything is obviously wrong. This is my first ever overclock. Thanks!

 

In BIOS

 

240

x 4

16 16

x10

100MHz

Disable

Auto

Startup

 

1.375

Auto

1.20

1.52

 

DRAM Configuration

200

Enable

3.0

4

8

3

11

16

2

3

2

3

3120

Auto

Enabled

 

Auto

0

Level 8

Level 1 (Reduce 50%)

8ns

Normal

6ns

256

Disable

16x

4x

Disable(4 bursts)

Share this post


Link to post
Share on other sites

You have another thread, "NF4 Expert - MemTest error!", but it ends with no further comments about if you figured out how to get your memory to pass. Did you? And if, maybe it would be good if you could end that thread with an explanaition about how - it will help others who search the forum - and PM a moderator so he could edit the thread to "solved".

Share this post


Link to post
Share on other sites

Yah thats what I was thinig put ur ram on a low devider and see if u pass then, try to pinpoint ur failure, im curently doin 2.7 with 1.55 volts ram on 166 devider, still not prime stable on core 0 but after sleeping at night and burning her in shes getting there..

Share this post


Link to post
Share on other sites

Yah thats what I was thinig put ur ram on a low devider and see if u pass then, try to pinpoint ur failure, im curently doin 2.7 with 1.55 volts ram on 166 devider, still not prime stable on core 0 but after sleeping at night and burning her in shes getting there..

 

A divider? I'm not really sure what that does but I'll try lol. Does anyone vote for just raising the voltage to 1.400? hehe

 

I should have stopped the stupid test at 12 hours. Then I would never have known about this crap.

Share this post


Link to post
Share on other sites

The cores might differ, so that's something we at times have to accept.

 

What kind of setup are you using for Prime95? Depending on settings it will more or less affect you RAM, so in some cases it's not a core failing, but RAM.

Share this post


Link to post
Share on other sites

What are you using for cooling and have you checked your TIM application? If you can keep the temps down I would go up on the volts.

 

Sorry, "TIM" application? If you mean SmartGuardian I'm keeping an eye on that, yes. As I said, I'm using a Thermalright XP-90 and a Vantec Tornado controlled by my PSU. Load temp at 1.375V was 45c.

Share this post


Link to post
Share on other sites

The cores might differ, so that's something we at times have to accept.

 

What kind of setup are you using for Prime95? Depending on settings it will more or less affect you RAM, so in some cases it's not a core failing, but RAM.

 

In Prime95 I'm running two instances from two different folders, each one set for affinity to one core. I have 2GB RAM, and I have the RAM usage on each instance set to 850MB.

 

EDIT: I am MemTest stable X 3 full test cycles (about 1.5 hours) on current timings and voltages.

Share this post


Link to post
Share on other sites

In Prime95 I'm running two instances from two different folders, each one set for affinity to one core. I have 2GB RAM, and I have the RAM usage on each instance set to 850MB.

Since you're now trying to test your CPU, as I can understand it, lower the RAM to be used, since this might cause it to fail. Drop it atleast to 750 MB or lower, you don't have to push your RAM right now, leave that to later.

 

Of course, you'll have some work to do next - if the RAM is the cause of failure - to figure out how to adjust RAM settings.

 

Edit: MemTest envoirement is quite different, since it does totally isolate the RAM to do certain operations. When your OS is running the situation is different.

 

Next: MemTest full cycles, ok, but if you choose test 5 for example, how does it work?

Share this post


Link to post
Share on other sites

I got another Expert and a x2 3800 Toledo this morning, managed to spend some time with it tonight - dual core is an interesting game. It's a CBBWE 0546UPMW, not an earth shattering stepping but quite intriguing. First real test for a 2GB G.Skill HZ kit too, so it's quite unfamiliar.

 

Can't wait to hit it with SS. Teh dielectric grease is agitating.

 

Duh, I forgot I was going to ask - what are safe max volts on stock HSF? 1.55 or 1.6? Does it vary between Manchester and Toledo cores?

Share this post


Link to post
Share on other sites

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...