[SOLVED] DPC ISSUE, micro stuttering
- Thread starter Calcifer
- Start date
RepairandRestore
Well-known member
Oh...
can you please remind me we stopped the integrated one from BIOS right?
can you please remind me we stopped the integrated one from BIOS right?
RepairandRestore
Well-known member
yes, i've already disabled it through UEFI. i've tried both official driver and uni xonar driver with LOW DPC install option. a thing that happens if i disable audio card and i use the computer without sound is that the driver nvidia causes high dpc. I think, i'm not sure though, i see cmudaxp.sys high due to nvidia driver. before reinstalling the OS i've tried to see if high dpc would have happen without it and i also unplugged all the usb devices and nvidia was the only one high.
thank you :)
thank you :)
OldGrayGary
Visiting Expert
- Jan 25, 2013
- 599
Hi again all
What I noticed most in your latest graphs is not any trouble with latency (which hasn't been bad since you tried the 364 version Nvidia drivers) --- it's the hard page fault counts that are high. You mention that the system as idle, and yet the hard page fault counts are far in the red. There is likely a process in the svchost module (I like to compare these to "bags of processes"... there can be fifteen or more processes all running like crazy inside of a svchost "container") that is causing the worst of the page fault overuse. You'll probably have to "catch it in action".
If you have time, you can try using Task Manager to try to peek at what process running inside an svchost container is causing the most hard page fault activity. Run Latency Monitor, and make note of the svchost module at the top of the hard page fault usage list. Then open Task Manager to the Processes view, and click the down arrow next to the name of that svchost container, and it should expand to show you the list of processes running inside that svchost. Note which processes look like suspects for high page fault counts -- likely the most active processes using the most memory. And post the information here.
[For example, the PID of the svchost container with the very high hard page fault counts was 1028 on that day -- the process IDs change from session to session ... but if you've just run Latency Monitor to look for high page counts, and find a current PID that's the highest, you can then search in Task Manager for that process]
Looks like you might want to close FireFox and Photos when gaming or watching video, since even when idle those two were high in hard page fault counts. You could look in, too, on your "Indexing Options" in Control Panel, and make sure that Search isn't over-indexing things you don't need it to.
You've managed to get latency under reasonable limits - let's see if we can get the hard page faults down too ... if we can, I think we have a decent chance of a no-stutter system.
What I noticed most in your latest graphs is not any trouble with latency (which hasn't been bad since you tried the 364 version Nvidia drivers) --- it's the hard page fault counts that are high. You mention that the system as idle, and yet the hard page fault counts are far in the red. There is likely a process in the svchost module (I like to compare these to "bags of processes"... there can be fifteen or more processes all running like crazy inside of a svchost "container") that is causing the worst of the page fault overuse. You'll probably have to "catch it in action".
If you have time, you can try using Task Manager to try to peek at what process running inside an svchost container is causing the most hard page fault activity. Run Latency Monitor, and make note of the svchost module at the top of the hard page fault usage list. Then open Task Manager to the Processes view, and click the down arrow next to the name of that svchost container, and it should expand to show you the list of processes running inside that svchost. Note which processes look like suspects for high page fault counts -- likely the most active processes using the most memory. And post the information here.
[For example, the PID of the svchost container with the very high hard page fault counts was 1028 on that day -- the process IDs change from session to session ... but if you've just run Latency Monitor to look for high page counts, and find a current PID that's the highest, you can then search in Task Manager for that process]
Looks like you might want to close FireFox and Photos when gaming or watching video, since even when idle those two were high in hard page fault counts. You could look in, too, on your "Indexing Options" in Control Panel, and make sure that Search isn't over-indexing things you don't need it to.
You've managed to get latency under reasonable limits - let's see if we can get the hard page faults down too ... if we can, I think we have a decent chance of a no-stutter system.
OldGrayGary
Visiting Expert
- Jan 25, 2013
- 599
Hi there
Sorry, I should have mentioned that it matters what view we are using. It shows as "svchost.exe" in the "Details" tab, but it shows as "Service Host...." in the "Processes" tab. The nice thing about the view in the "Processes" tab is that there is a short description following "Service Host", such as Local Service, Local System, Network Service, etc...
By the way, there is a new Cumulative Update for Windows 10 available today. It takes a while to install - but it might be worth installing right away, in case any of the fixes included help with the audio/visual stuttering problems. I put a brief list of the fixes in a thread here in our Windows 10 forum....
Who knows, it might help!
Sorry, I should have mentioned that it matters what view we are using. It shows as "svchost.exe" in the "Details" tab, but it shows as "Service Host...." in the "Processes" tab. The nice thing about the view in the "Processes" tab is that there is a short description following "Service Host", such as Local Service, Local System, Network Service, etc...
By the way, there is a new Cumulative Update for Windows 10 available today. It takes a while to install - but it might be worth installing right away, in case any of the fixes included help with the audio/visual stuttering problems. I put a brief list of the fixes in a thread here in our Windows 10 forum....
Who knows, it might help!
OldGrayGary
Visiting Expert
- Jan 25, 2013
- 599
Hi again Calcifer
The high page fault counts will show up in Latency Monitor (or in a Performance Monitor cpu trace that can be viewed with Windows Performance Analyzer). The new 6.50 Windows-10 compatible Latency Monitor from Resplendence is a decent tool - you'd sort the list for page fault results by process, and then look at the top two or three processes in the list. (You can do the same using Performance Monitor & WPA). Once you have a list of top offenders, that's when you start looking in Task Manager to see what processes are running inside the "Service Host" mentioned in the page fault graphs.
If you'd like to try to simply "catch the devil in action" - you could download and install "Process Monitor" from Microsoft's SysInternals site. Process Monitor is a lot like Task Manager, but with more details. And since its from Microsoft, you can trust it to get along with Windows 10, and to produce reasonably accurate results.
Process Monitor
See what turns up
The high page fault counts will show up in Latency Monitor (or in a Performance Monitor cpu trace that can be viewed with Windows Performance Analyzer). The new 6.50 Windows-10 compatible Latency Monitor from Resplendence is a decent tool - you'd sort the list for page fault results by process, and then look at the top two or three processes in the list. (You can do the same using Performance Monitor & WPA). Once you have a list of top offenders, that's when you start looking in Task Manager to see what processes are running inside the "Service Host" mentioned in the page fault graphs.
If you'd like to try to simply "catch the devil in action" - you could download and install "Process Monitor" from Microsoft's SysInternals site. Process Monitor is a lot like Task Manager, but with more details. And since its from Microsoft, you can trust it to get along with Windows 10, and to produce reasonably accurate results.
Process Monitor
See what turns up
Hi again OldGrayGary,
as always thanks for keep helping me out.
i was going to do what you said so i started latencymon before having lunch and see at the end the svchost process but... now it's changed.. again.
here's the log and screenshots :
could the last update have change something ?
this report like the last time it's during idle time and it changes at all during load session with cmudaxp.sys and nvlddmkm.sys at the top.
this is confusing because i don't know if idle session reports ar related with load session reports.
once again, thank you.
as always thanks for keep helping me out.
i was going to do what you said so i started latencymon before having lunch and see at the end the svchost process but... now it's changed.. again.
here's the log and screenshots :
Read More:
could the last update have change something ?
this report like the last time it's during idle time and it changes at all during load session with cmudaxp.sys and nvlddmkm.sys at the top.
this is confusing because i don't know if idle session reports ar related with load session reports.
once again, thank you.
OldGrayGary
Visiting Expert
- Jan 25, 2013
- 599
Hi again
I'd be interested to see the graphs when under load.
The fact that this last set of graphs was created when you system was supposed to be "idle", but yet shows the very high "hard pagefaults" from the Microsoft Photos app and the Runtime Broker, is a rather telling clue: you definitely want to close Photos and not allow it to gobble up resources unnecessarily. https://www.reddit.com/r/Surface/comments/3zxwnt/microsoft_photos_app_cpu_hog/
You also might want to check a setting for Windows Updates, which might cause unnecessary background traffic on your network (and interrupt your own activities if heavy enough)
1) Click on the Windows Start Menu icon
2) Select Settings
3) Select Update & Security
4) Select Windows Update
5) Select Advanced Options
6) Select Choose How Updates Are Delivered
7) Make sure that the option "Updates from more than one place" is turned OFF.
You could also be experiencing odd behavior from a Notifications setting. Some users have reported that the background automatic notifications go a bit bonkers, and cause spikes in cpu usage (while no "noticeable" benefit of all the activity can be found -- [get the pun :)] )... The fix is a simple little tweak, and completely harmless (beneficial in most respects, really).
Turn off "tips" in Notifications
1) Click on the Windows Start Menu icon
2) Select Settings
3) Select System
4) Select Notifications and Actions
5) Turn "Show me tips about Windows" to OFF
If you aren't using any particular alarms, and don't use Windows Mail, or News, or other notifying apps ... you can turn off those notifications as well. The switch toggles OFF/On with just a click.
See if any of this helps.
I'd be interested to see the graphs when under load.
The fact that this last set of graphs was created when you system was supposed to be "idle", but yet shows the very high "hard pagefaults" from the Microsoft Photos app and the Runtime Broker, is a rather telling clue: you definitely want to close Photos and not allow it to gobble up resources unnecessarily. https://www.reddit.com/r/Surface/comments/3zxwnt/microsoft_photos_app_cpu_hog/
You also might want to check a setting for Windows Updates, which might cause unnecessary background traffic on your network (and interrupt your own activities if heavy enough)
1) Click on the Windows Start Menu icon
2) Select Settings
3) Select Update & Security
4) Select Windows Update
5) Select Advanced Options
6) Select Choose How Updates Are Delivered
7) Make sure that the option "Updates from more than one place" is turned OFF.
You could also be experiencing odd behavior from a Notifications setting. Some users have reported that the background automatic notifications go a bit bonkers, and cause spikes in cpu usage (while no "noticeable" benefit of all the activity can be found -- [get the pun :)] )... The fix is a simple little tweak, and completely harmless (beneficial in most respects, really).
Turn off "tips" in Notifications
1) Click on the Windows Start Menu icon
2) Select Settings
3) Select System
4) Select Notifications and Actions
5) Turn "Show me tips about Windows" to OFF
If you aren't using any particular alarms, and don't use Windows Mail, or News, or other notifying apps ... you can turn off those notifications as well. The switch toggles OFF/On with just a click.
See if any of this helps.
OldGrayGary
Visiting Expert
- Jan 25, 2013
- 599
Hi again Calcifer
We've had a lot of issues similar to yours in the forums lately... so, in addition to the three steps just mentioned (turn off Photos, Windows Update settings change, and Notification settings change) .... I have a few more steps to recommend (since there doesn't seem to be any guarantee for a single fix)...
Storport.sys/iastor.sys & Intel Rapid Storage Technology
I've actually recommended this for a few other troubled systems in the past, but it's been a while, so I'm going to link to a how-to (rather than trying to remember each exact step). Basically, you move from using the Intel-specific Rapid Storage Technology drivers, to using generic drivers.
Uninstalling the Intel(R) Rapid Storage Technology driver - Microsoft Community
Generic USB 3.0 driver rather than Asus/Asrock provided USB 3.0 driver
If the current USB 3.0 driver is the Asus/Asrock version (usually downloaded from their website, or installed from a "driver CD" that comes with the motherboard), try using a generic Intel USB 3.0 driver. You can try uninstalling the current driver using Device Manager, and then visiting Windows update (probably before you restart [so that you don't end up reinstalling the same Asus-version driver]) and let Windows Update provide whatever USB 3.0 driver it has in its library.
High Interrupt count for the nvlddmkm.sys Nvidia driver
I've been looking into several threads with this driver experiencing high interrupt counts - and your computer is one of those showing this effect. I think, but I'm not sure yet, that the driver has to do with power management for the Nvidia card. There have been several strategies used to varying success mentioned in tech forums, here are a few of the ideas:
1) If you haven't already, set your Energy Plan to "Maximum Performance" in Control Panel's "Power Options" app.
2) If that step doesn't help, you can try using an old app to lower the Nvidia power management settings, called Nvidia "PowerMizer". This used to help a few users who had Windows 7 - some Windows 10 users report that it helps in 10 also ... [though some report that sometimes updates arrive, and then they have to re-set the power settings again using PowerMizer]. The trick is to lower the settings, as mentioned in the thread (scroll down to the reply by Douglas Wilcox):
Screen going black but still running - General Hardware - Laptop - Dell Community
3) If neither of those steps help, you can try an alternate/opposite attempt --- use a few Registry tweaks to disable the power management of nvlddmkm.sys ... this will only work if the Windows 10 Nvidia Registry settings are in the same places as the Windows 7 settings...
SOLVED! Dropouts, Cracks, Pops on Windows 7 and NVIDIA GFX CARD! | NI User Forum
A few systems are trying these steps --- let us know if you have any success...
We've had a lot of issues similar to yours in the forums lately... so, in addition to the three steps just mentioned (turn off Photos, Windows Update settings change, and Notification settings change) .... I have a few more steps to recommend (since there doesn't seem to be any guarantee for a single fix)...
Storport.sys/iastor.sys & Intel Rapid Storage Technology
I've actually recommended this for a few other troubled systems in the past, but it's been a while, so I'm going to link to a how-to (rather than trying to remember each exact step). Basically, you move from using the Intel-specific Rapid Storage Technology drivers, to using generic drivers.
Uninstalling the Intel(R) Rapid Storage Technology driver - Microsoft Community
Generic USB 3.0 driver rather than Asus/Asrock provided USB 3.0 driver
If the current USB 3.0 driver is the Asus/Asrock version (usually downloaded from their website, or installed from a "driver CD" that comes with the motherboard), try using a generic Intel USB 3.0 driver. You can try uninstalling the current driver using Device Manager, and then visiting Windows update (probably before you restart [so that you don't end up reinstalling the same Asus-version driver]) and let Windows Update provide whatever USB 3.0 driver it has in its library.
High Interrupt count for the nvlddmkm.sys Nvidia driver
I've been looking into several threads with this driver experiencing high interrupt counts - and your computer is one of those showing this effect. I think, but I'm not sure yet, that the driver has to do with power management for the Nvidia card. There have been several strategies used to varying success mentioned in tech forums, here are a few of the ideas:
1) If you haven't already, set your Energy Plan to "Maximum Performance" in Control Panel's "Power Options" app.
2) If that step doesn't help, you can try using an old app to lower the Nvidia power management settings, called Nvidia "PowerMizer". This used to help a few users who had Windows 7 - some Windows 10 users report that it helps in 10 also ... [though some report that sometimes updates arrive, and then they have to re-set the power settings again using PowerMizer]. The trick is to lower the settings, as mentioned in the thread (scroll down to the reply by Douglas Wilcox):
Screen going black but still running - General Hardware - Laptop - Dell Community
3) If neither of those steps help, you can try an alternate/opposite attempt --- use a few Registry tweaks to disable the power management of nvlddmkm.sys ... this will only work if the Windows 10 Nvidia Registry settings are in the same places as the Windows 7 settings...
SOLVED! Dropouts, Cracks, Pops on Windows 7 and NVIDIA GFX CARD! | NI User Forum
A few systems are trying these steps --- let us know if you have any success...
hi again OldGrayGary
i don't have intel rapid storage installed because as you adviced i kept only microsoft driver in this new installation and the same is for the USB driver.
the only drivers i downloaded are the updated one from nvidia and the uni xonar for the audio card because windows didn't provide them.
unfortunately i already have Prefere Maximum Performance on nvidia panel. I'll let you know about the point 2 and 3 .
sorry for the short answer but i'm a bit despondent about this problem. this issue is taking so much time :)
oh by the way i uninstalled windows.photo but after that pictures could be only open through paint.
as always thank you for your kind support.
update :
so i tried set the paging file as you adviced, i think i had some micro stutter during game session but i'm not sure if the cause could be latency mon itself.
here's the screenshots and logs :
i'm not sure if the highest pagefault resolution time could be caused by me doing alt+tab to put the game down and see how was going latency mon.
now, how can paging file be related to a driver issue ? i'm more confused now.
furthremore my ram is 8gb ( 2 x 4 ) 1600 hz but in cpu-z i see this :
i also checked on UEFI and DRAM FREQUENCY it's set to 1333 (1:5)
could this be the problem since i have 1600 frequency ram ?
thank you
update :
i changed it in the UEFI to 1600 and now cpu-z show this :
so i tried set the paging file as you adviced, i think i had some micro stutter during game session but i'm not sure if the cause could be latency mon itself.
here's the screenshots and logs :
Read More:
i'm not sure if the highest pagefault resolution time could be caused by me doing alt+tab to put the game down and see how was going latency mon.
now, how can paging file be related to a driver issue ? i'm more confused now.
furthremore my ram is 8gb ( 2 x 4 ) 1600 hz but in cpu-z i see this :
i also checked on UEFI and DRAM FREQUENCY it's set to 1333 (1:5)
could this be the problem since i have 1600 frequency ram ?
thank you
update :
i changed it in the UEFI to 1600 and now cpu-z show this :
UPDATE ( sorry... i can't edit previous reply , don't know why )
i installed an older uni xonar driver version ( more stable they say ) and i disabled dolby digital. as you can see turning it off gave me low latency.
i had micro stuttering but the report it's strange.
here's the log and screenshot :
ah by the way, i'm almost sure the highest hardpage file reported resolution time goes high when i alt tab because it's the 2nd time i see it while i put the game down to see latency mon.
i didn't turned off paging file because i'm not sure 8gb ram are enough and i wanted to try your first advice. by the way, i restarted but the OS didn't asked me to after i changed paging file. is it normal ?
i noticed a reduced amount of hardpage fault doing that.
ah about the sfcfix you gave some days ago that worked, a folder was created and i deleted it and nothing happened ( wasn't sure if i could or not ) :) i did another sfc /scannow for security and it was fine.
sorry for all these updates.
thank you again.
i installed an older uni xonar driver version ( more stable they say ) and i disabled dolby digital. as you can see turning it off gave me low latency.
i had micro stuttering but the report it's strange.
here's the log and screenshot :
Read More:
ah by the way, i'm almost sure the highest hardpage file reported resolution time goes high when i alt tab because it's the 2nd time i see it while i put the game down to see latency mon.
i didn't turned off paging file because i'm not sure 8gb ram are enough and i wanted to try your first advice. by the way, i restarted but the OS didn't asked me to after i changed paging file. is it normal ?
i noticed a reduced amount of hardpage fault doing that.
ah about the sfcfix you gave some days ago that worked, a folder was created and i deleted it and nothing happened ( wasn't sure if i could or not ) :) i did another sfc /scannow for security and it was fine.
sorry for all these updates.
thank you again.
OldGrayGary
Visiting Expert
- Jan 25, 2013
- 599
Hi again
Can you double-check that you are using the 6.50 version of Latency Monitor? [If it turns out you have an older version, uninstall it, and download the newer one. The latest is available directly from resplendence.com
The way CPU-Z presents its results ... can be confusing, since the information in your computer travels at "multiples" of the standard speeds across certain connections. DDR memory (DDR, DDR2, DDR3, DDR4...) is "Double Data Rate" memory: this type of memory delivers two chunks of data per clock cycle (as compared to just one for the older SDRAM type of memory ... you can think of the S as "single" compared to "double" [though it actually stands for synchronous]). So, the memory in your system when set to DDR3-1333 travels in double "chunks" at 666.6. / and when set to DDR3-1600, data travels in double "chunks" at 800. So long as your motherboard and your memory support the faster setting, you might see a little improvement since you've boosted the speed to it's "rated" speed.
The numbers look better in your first of your latest two graphs, apparently League of Legends doesn't generate nearly as high a hard pagefault count as Battle.net -- which dragged along quite a lot... higher interrupts, especially. The results will likely be different for different games - hopefully we can find a way to have all of them play without audio troubles.
-- I'm especially interested to hear about your version of Latency Monitor ... because of the odd measurements reported for your computer's cpu speed. It might be a glitch in Latency Monitor - I'll have to take a look and see ....
_________________________
I think the pagefile should be OK with the generous custom-set range. Let's leave it like that for now.
Just to double-check the SATA/AHCI side of things: can you take another quick peek in Device Manager, click on the right-facing-arrow in front of IDE ATA/ATAPI Controllers, right-click Standard SATA AHCI Controller, select Properties, select the Driver tab, select Driver Details, and verify that the driver is listed as "storahci.sys"?
.............. and thanks for sending all the graphs and info!
Can you double-check that you are using the 6.50 version of Latency Monitor? [If it turns out you have an older version, uninstall it, and download the newer one. The latest is available directly from resplendence.com
The way CPU-Z presents its results ... can be confusing, since the information in your computer travels at "multiples" of the standard speeds across certain connections. DDR memory (DDR, DDR2, DDR3, DDR4...) is "Double Data Rate" memory: this type of memory delivers two chunks of data per clock cycle (as compared to just one for the older SDRAM type of memory ... you can think of the S as "single" compared to "double" [though it actually stands for synchronous]). So, the memory in your system when set to DDR3-1333 travels in double "chunks" at 666.6. / and when set to DDR3-1600, data travels in double "chunks" at 800. So long as your motherboard and your memory support the faster setting, you might see a little improvement since you've boosted the speed to it's "rated" speed.
The numbers look better in your first of your latest two graphs, apparently League of Legends doesn't generate nearly as high a hard pagefault count as Battle.net -- which dragged along quite a lot... higher interrupts, especially. The results will likely be different for different games - hopefully we can find a way to have all of them play without audio troubles.
-- I'm especially interested to hear about your version of Latency Monitor ... because of the odd measurements reported for your computer's cpu speed. It might be a glitch in Latency Monitor - I'll have to take a look and see ....
_________________________
I think the pagefile should be OK with the generous custom-set range. Let's leave it like that for now.
Just to double-check the SATA/AHCI side of things: can you take another quick peek in Device Manager, click on the right-facing-arrow in front of IDE ATA/ATAPI Controllers, right-click Standard SATA AHCI Controller, select Properties, select the Driver tab, select Driver Details, and verify that the driver is listed as "storahci.sys"?
.............. and thanks for sending all the graphs and info!
hi OldGrayGray
latencymon version is 6.5 ( you can see it on the top of the screenshots :) )
here's the SATA/ACHI driver version :
furthermore i see a strange icon on some devices, i did a circle on them. what does that mean ?
anyway, i'm not 100% sure but, i think i don't have any audio problem at the moment, just micro stuttering resulting in an half second screen freeze .
thank you again for your support.
latencymon version is 6.5 ( you can see it on the top of the screenshots :) )
here's the SATA/ACHI driver version :
furthermore i see a strange icon on some devices, i did a circle on them. what does that mean ?
anyway, i'm not 100% sure but, i think i don't have any audio problem at the moment, just micro stuttering resulting in an half second screen freeze .
thank you again for your support.
update :
ok i had normal use before having lunch and latencymon was fine, listening music, going on web, watching videos on computer. all fine. then i turned off the monitor and left the computer on, so in idle. i put on the monitor and here's the report :
unfortunately when i open the taskmanager i can't see any host service with high ram usage and the computer runs fine when i use it. i wonder what can happens when i don't use it for more than 10 minutes ( when the monitor get disabled, since i removed only the computer suspension/hybernation and not the monitor one). Could the high nvidia driver execution time be caused by the monitor getting enabled again ?
at this point, i'm almost sure to say that the problems occur only when gaming and when the computer is idle, even though i don't know if they are related.
thank you.
ok i had normal use before having lunch and latencymon was fine, listening music, going on web, watching videos on computer. all fine. then i turned off the monitor and left the computer on, so in idle. i put on the monitor and here's the report :
Read More:
unfortunately when i open the taskmanager i can't see any host service with high ram usage and the computer runs fine when i use it. i wonder what can happens when i don't use it for more than 10 minutes ( when the monitor get disabled, since i removed only the computer suspension/hybernation and not the monitor one). Could the high nvidia driver execution time be caused by the monitor getting enabled again ?
at this point, i'm almost sure to say that the problems occur only when gaming and when the computer is idle, even though i don't know if they are related.
thank you.
OldGrayGary
Visiting Expert
- Jan 25, 2013
- 599
Hi again
Latest Graphs
Things are looking quite a bit better in your latest graphs. The interrupts and latency counts are much better. The high pagefault is still a bit of a mystery, but once again it seems to be something running inside a "Service Host" (or "svchost.exe"). It could be almost anything, and it might well be a simple maintenance item (like Superfetch, or Indexing). Some of these maintenance items you can change (schedule manually, instead of automatically, for example).
To have a look at what might be running inside a particular "Service Host", you can try a free utility from Microsoft's SysInternals called "Process Explorer". It is a lot like the "Processes" and "Details" tabs of Task Manager all on one screen. It shows the numbers for each process running inside a Service Host, for example - while Task Manager's "Processes" tab only shows the totals for all the processes lumped together. If you let Process Monitor run while you are testing for micro-stutters, when they start happening, pop up the Process Monitor screen, and see if you can catch the culprit in action. I recommend not running Latency Monitor and Process Explorer at the same time .... it might skew the numbers...
Here's a link to Process Explorer over at Microsoft's site:
Process Explorer
______________________________
Strange icons in Device Manager
The new icons you see in Device Manager are not because you have an error or warning condition: they are just a new design in Windows 10 to show a PCI device. The icon is supposed to look like a PCIe card. I have them in my Device Manager screen too:
..... glad that things are a bit better. Have a look in on what's running while micro-stutters happen with Process Explorer, and see if anything shows up.
Cheers
Latest Graphs
Things are looking quite a bit better in your latest graphs. The interrupts and latency counts are much better. The high pagefault is still a bit of a mystery, but once again it seems to be something running inside a "Service Host" (or "svchost.exe"). It could be almost anything, and it might well be a simple maintenance item (like Superfetch, or Indexing). Some of these maintenance items you can change (schedule manually, instead of automatically, for example).
To have a look at what might be running inside a particular "Service Host", you can try a free utility from Microsoft's SysInternals called "Process Explorer". It is a lot like the "Processes" and "Details" tabs of Task Manager all on one screen. It shows the numbers for each process running inside a Service Host, for example - while Task Manager's "Processes" tab only shows the totals for all the processes lumped together. If you let Process Monitor run while you are testing for micro-stutters, when they start happening, pop up the Process Monitor screen, and see if you can catch the culprit in action. I recommend not running Latency Monitor and Process Explorer at the same time .... it might skew the numbers...
Here's a link to Process Explorer over at Microsoft's site:
Process Explorer
______________________________
Strange icons in Device Manager
The new icons you see in Device Manager are not because you have an error or warning condition: they are just a new design in Windows 10 to show a PCI device. The icon is supposed to look like a PCIe card. I have them in my Device Manager screen too:
..... glad that things are a bit better. Have a look in on what's running while micro-stutters happen with Process Explorer, and see if anything shows up.
Cheers
hi again,
ok i'll let you know if i'll find something. seems to be stuck with the issue at the moment unfortunately.
p.s. i watched "the wind rises". honestly, it's not my favorite movie of the studio ghibli. *MAYBE A SPOILER ! * it's sad ! the main character seems a bit selfish, wasn't he ? he should have stay with her ! *
ok i'll let you know if i'll find something. seems to be stuck with the issue at the moment unfortunately.
p.s. i watched "the wind rises". honestly, it's not my favorite movie of the studio ghibli. *MAYBE A SPOILER ! * it's sad ! the main character seems a bit selfish, wasn't he ? he should have stay with her ! *
Has Sysnative Forums helped you? Please consider donating to help us support the site!