Jump to content
AIDA64 Discussion Forum
Humajum

DPC Latency

Recommended Posts

Is there anyway to lower the DPC latency while running AIDA64? I get periodic spikes in DPC latency while running AIDA64 to monitor temps on my G19. Sometimes my audio playback will skip because of this; any help would be much appreciated :)

Share this post


Link to post
Share on other sites

Please let us know more about your system configuration:

1) Motherboard model

2) CPU type

3) Video card (or cards) model

4) Number of internal or external hard disk drives or SSD

5) Any special sensor devices connected

6) Windows variant, kernel type (32-bit or 64-bit)

Thanks,

Fiery

Share this post


Link to post
Share on other sites

1) Gigabyte EX58-UD5

2) i7 930

3) GTX 480

4) 2x Intel SSDs in RAID0, 3x1TB Caviar Green in RAID0 (ICH10R)

5) No sensor devices

6) Win7 Pro x64

I've attached a picture of the dpc latency, it spikes fairly regularly while leaving AIDA64 running.

post-189-035828900 1293589522_thumb.jpg

Share this post


Link to post
Share on other sites

i believe i got also some latency issues, last day i was playback some music and got sound stutter regularly. I'm not sure aid64 was launched or not but it may be. Next time I'll check with latency mon to be sure.

Share this post


Link to post
Share on other sites

In case the latency spikes are approx. 20 seconds away from each other -- as it seems --, then I suppose it's caused by the SMART detection layer of AIDA64, which is used to measure HDD/SSD temperature. AIDA64 now uses a special driver call to pass through the Intel Rapid Storage RAID drivers, in order to reach the RAID member drives. I guess Intel drivers will need some more iterations to mature :)

Share this post


Link to post
Share on other sites

Thanks for all the replies. LatencyMon reports everything as ok but closing it and running DPC latency checker will show periodic spikes. May just be the Intel RST drivers as Fiery stated, I recently moved to the new v10 that was released on 12/15. Thanks again.

Share this post


Link to post
Share on other sites

Is there a difference between Latency Monitor and DPC Latency Checker? Just curious. I use DPC Latency Checker ever since you guys mentioned it here, I didn't know about it. No drop-outs for me on Windows 7 Pro 64.

Share this post


Link to post
Share on other sites

latency monitor actually displays the driver at fault, and has extended DPC information including a DPC amount counter.

Excessive DPC activity is just as problematic as high dpc spikes. the driver used is also more recently updated then dpc latency checker.

Share this post


Link to post
Share on other sites

Well, dunno if this is Gigabyte releated, but got here GA-X58A-UD7 Rev. 2 board and also clearly DPC latency caused by AIDA64.

The Program runs for 60 seconds even and DPC jumps to 60000-135000 every HW scan AIDA64 makes. Tested v1.20 and v1.50 and both AIDA versions causes this. Same does not happen, if I run HWMonitor or any of the other tools and this doesn't seem to happen when I run Everest 5.50.

Why I think this is Gigabyte releated is because this does not happen on ASUS-P5E (x38) or ASUS-P6T (X58) boards at all. Only way to stop the DPC jumps is to stop kernel driver from AIDA64 even disabling all other checkmarks on stability section doesn't help.

Share this post


Link to post
Share on other sites

gigabyte has a history of implementing sensors and hardware in a weird manner.

Yeah, understand, but don't you think it's just a little bit weird this doesn't happen on Everest ? Most also be something really small like some ACPI check but the checks are already unchecked while installing AIDA64, however, when I visit the area of information seems the jumps come more regularly there while going through information.

heck, if I wouldn't know better I could read through several topics on forums like:

and I bet these are all caused by same looped check somewhere.

Share this post


Link to post
Share on other sites

I too have the Gigabyte-X58A-UD7 board, but it's the revision 1 version. I run DPC Latency with no spikes or problems. I don't use Gigabyte's Easy Tune software, that's a good point to check if you have that installed.

Share this post


Link to post
Share on other sites

do you have easy tune installed? i believe that installs a service which maintains a driver running in the background.

it will help fiery to know if that plays a part.

Nope, no easytune nor any other releated tools installed or extra drivers (checked by 'autoruns') on background. No OC.

Share this post


Link to post
Share on other sites

Yeah, understand, but don't you think it's just a little bit weird this doesn't happen on Everest ? Most also be something really small like some ACPI check but the checks are already unchecked while installing AIDA64, however, when I visit the area of information seems the jumps come more regularly there while going through information.

One of the differences between AIDA64 and the last release of EVEREST is that AIDA64 supports a lot of new hardware (e.g. Sandy Bridge processors, SSD drives, Radeon HD 6000 family, GeForce GTX 500 family, etc) that EVEREST doesn't. And that effectively means that -- unlike EVEREST -- AIDA64 is able to monitor those new hardware, hence it reads more information from new hardware.

Share this post


Link to post
Share on other sites

One of the differences between AIDA64 and the last release of EVEREST is that AIDA64 supports a lot of new hardware (e.g. Sandy Bridge processors, SSD drives, Radeon HD 6000 family, GeForce GTX 500 family, etc) that EVEREST doesn't. And that effectively means that -- unlike EVEREST -- AIDA64 is able to monitor those new hardware, hence it reads more information from new hardware.

Yeah, that goes for PR, but was referring what exactly was causing the issue. ;)

Seems if I uncheck all the stability section except the kernel mode driver the DPC latency lowers to ~20-30K instead of 60-130K jumps. So most be multiple issues there. Further I go down from Stability section more specifically SMBus section checked we yet again get the 60-13K jumps. However, it still jumps quite a bit even on Kernel Mode Driver enabled and not else.

Share this post


Link to post
Share on other sites

gigabyte has a history of implementing sensors and hardware in a weird manner.

well.what about on intel motherboards?one of my friend is the same config as mine just with an intel board.he is also having the same issue as mine.so,how you are going to explain that.that intel also has sensor issue?

i just wish my problem is solved by the next release or something like patch/script

Share this post


Link to post
Share on other sites

then you don't read enough websites, the switching between performance states can trigger a dpc spike, and HPET itself increases its average level during use.

another cause of high DPC is the current intel raid drivers.

Share this post


Link to post
Share on other sites

might not even be AIDA,

HPET and Intel's enhanced C states also cause dpc issues.

That's actually quite curious idea. Wondering since I am using 64-bit HPET by BIOS could that actually be causing this need to test the theory. As for enhanced C states well that quite long reach I'd say considering these are timed events and updates happening every 3 seconds pretty much makes sure no event would power save in this time (well other that something like CPU).

But really need to check on HPET state and see powercfg.exe for raise hardware wake-up events and see where the lag starts to appear. Although, system is at full performance mode and even HD internal states are disabled. Including all CPU power saving states and all PCI-E power save states except GPU power saving states.

another cause of high DPC is the current intel raid drivers.

What's wrong with 10.1.0.1008 ? I know there's quite an bug in these or well actually it's not the Intel drivers, but their ICH10R Option ROM has bug which when caching is dropped drops also Disk internal read caching which was pretty hilarious to try to solve over Gigabyte support I might add since neither Intel Support nor Gigabyte support couldn't solve the Option ROM to only drop write caching when asked and not the read caching while at it. ;)

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.

Loading...

×
×
  • Create New...