CPU overclocking "bad_pool_caller"

[Mini0510]

I use realbench to stress test. When I stop it after a few min, it gives me this error. The error is gone after increase vcore by maybe 0.03-0.04 volts.

The weird thing is, I can run prime95 for even longer and stop it without this blue screen "bad_pool_caller" error.

 

Anyone have any ideas or reasoning? 

 

 

EDIT: Loadline calibration is set to 2nd highest level. Barely any vdrops.

[Electronics Wizardy]

Your system isn't stable when it blue screens. You only get the bluescreen running realbench as it uses different instructions that are stable at different speeds compared to prime 95.

[Mini0510]

1 minute ago, Electronics Wizardy said:

Your system isn't stable when it blue screens. You only get the bluescreen running realbench as it uses different instructions that are stable at different speeds compared to prime 95.

But it only happens when I stop realbench though. I've tested this on x79 and LGA 1155 platforms.

I guess there's not much of a point using prime95 over realbench then

[SageOfSpice]

 It could be vdroop from the processor immediately coming from a 100% load to idle. I don't know if load line calibration would fix that though, so you might just be stuck with running a slightly higher voltage.

[Mini0510]

2 minutes ago, SageOfSpice said:

 It could be vdroop from the processor immediately coming from a 100% load to idle. I don't know if load line calibration would fix that though, so you might just be stuck with running a slightly higher voltage.

Loadline calibration is set to 2nd highest level.
well, same thing happens through quite a few different combinations of CPUs and motherboards. I'm wondering if other people have encountered the same thing.

It's always when I stop realbench after running for a few min (time doesn't matter), "bad_pool_caller" blue screen. And I'm using windows 10.
But I guess it's due to not enough vcore since when I open like 20 chrome pages from my bookmarks one after another, the computer freezes because CPU and memory gets utilized alot

 

That's like my short and dirty test, opening tons of chrome pages from my bookmarks one after another 

[SageOfSpice]

4 minutes ago, Mini0510 said:

Loadline calibration is set to 2nd highest level.
well, same thing happens through quite a few different combinations of CPUs and motherboards. I'm wondering if other people have encountered the same thing.

It's always when I stop realbench after running for a few min (time doesn't matter), "bad_pool_caller" blue screen. And I'm using windows 10.
But I guess it's due to not enough vcore since when I open like 20 chrome pages from my bookmarks one after another, the computer freezes because CPU and memory gets utilized alot

 

That's like my short and dirty test, opening tons of chrome pages from my bookmarks one after another 

That's why I'm saying vdroop. 

It's got enough power to run, and it's got enough power to not throw errors at full load, but it doesn't have quite enough power to counteract the vdroop when it goes from it's loaded state to it's idle state, so power fluctuations.

 

A higher quality motherboard and possibly PSU might be able to alleviate that to some degree, but bumping the voltage up a bit isn't going to hurt anything.

[Mini0510]

1 hour ago, SageOfSpice said:

That's why I'm saying vdroop. 

It's got enough power to run, and it's got enough power to not throw errors at full load, but it doesn't have quite enough power to counteract the vdroop when it goes from it's loaded state to it's idle state, so power fluctuations.

 

A higher quality motherboard and possibly PSU might be able to alleviate that to some degree, but bumping the voltage up a bit isn't going to hurt anything.

ok but one of the motherboard is an asus maximus IV extreme. I don't think it's the motherboard issue.

PSU is 1000W. I was testing x79 so had to use that. on the LGA 1155, i ran it with a Corsair 750W PSU though. But should still be fine.

so the increase in vcore by 0.04v is justified? not much else I can do?

[SageOfSpice]

1 minute ago, Mini0510 said:

ok but one of the motherboard is an asus maximus IV extreme. I don't think it's the motherboard issue.

PSU is 1000W. I was testing x79 so had to use that. on the LGA 1155, i ran it with a Corsair 750W PSU though. But should still be fine.

so the increase in vcore by 0.04v is justified? not much else I can do?

I'm not saying your motherboard and PSU are BAD, just that they might not handle voltage at that scale as well as another board might.

 

If it works, it works man. 0.04v is virtually nothing.