Squall Leonhart

and if you're still checking back now and then Arctucas, 11.5.0.1153beta is now available

Ah, something that wouldn't cause a problem for everyone then? you're on the ball this week Fiery, i've never had a faster response xD i just hope intel have fixed their uninstall script, the previous version of 11.5 installed the iastorf (Filter) as a lowerfilter and failed to remove it on uninstall leading to fun times getting windows to work if you delete the file before removing the lowerfilter value

It may be the Win8CP Ndis bug, alot of Consumer preview testers are seeing the NDIS driver consume alot of cpu and cause extremely high dpc latency which could be delaying the aida64 cache test.

awesome

It should not require fixing 3.2 is not out just yet, but so far 11.5.0.1153beta tests have been positive

Fiery, do you have a Intel contact you should talk to about this?, similar behavior is found on Sandy and Ivy Bridge, but not Sandy Bridge E.

Open services.msc and set Task Scheduler to automatic

On these cpu's, it appears C3 is required to get the best results in the aida benchmark however C3 and C6 come at the expense higher DPC latency, and i also find windows to be smoother with them off :\

AIDA64 does have a MemBW test in the debug menu, but i don't recommend running it since it is almost impossible to cleanly stop.

it has nothing to do with AIDA64, the problem is entirely a hardware fault. Amazing how everyone else with a non evga gtx 680 can open aida64 just fine and not have any problems.

disable the smbus via nvidia i2c/driver setting and let aida64 access it via its own method. see if the issue persists, you actually get less data from the smbus access on certain cards (in the case of my 275 i lose fan rpm's and gpu VCC) if you have Precision X running alongside AIDA64, its advised to close one or the other. If precision X is accessing voltage adjustment via a HWMon module, then there are known issues between HWMon access mutexs and the mutex aida64 uses. (i've seen black screens and the gpu fan stopping)

these thinkpads seem to use an unusual acpi driver.

working in 10, 11, 12, 13 and 14 (i tried release through nightly)

some other times you can find a instability only by using light load.

in which it works just fine.....

no, but C1E is the cause of clock instabilities while Speedstep isn't.

i have, just never had to with aida64.

use the task scheduler instead.

look in administrative tools > task scheduler. its possible the delayed start isn't long enough and the taskhost is being stalled by another app/service you have starting at login.

C1E is enabled yet C state and speedstep are disabled.... am i the only one that finds the sillyness in this?

11.5 should return smart data if the device class is changed from SCSIAdapter to hdc SMART isn't the only thing busted on 11.5 though, i couldn't get perfectdisk to launch correctly with this driver either.

i see, well Dedicated and Dynamic memory appear to be used video memory, and probably don't seperate the commited from the shared video memory usage.

Where are these found in the main window now? i looked under DirectX and GPU and couldn't see them. never mind, they were right up the top of the direct3d list >.< so whats the difference with the new Dedicated and dynamic video memory?

afaik, they either check via DDraw, (though this has issues when reporting back more than 1GB) or use an undocumented driver method

Fiery, the IRST 11.5.0.1009a and IRST Enterprise drivers install as Class=SCSIAdapter(Storage Controllers), as opposed to Class=hdc (IDE/ATAPI) and SMART is not available via SCSIAdapter drivers for some reason. i expect this is part of the problem on x79