What's new

Do ALL SP2s suffer from hi CPU utilization when idle due to faulty audio drivers?

Have you done anything like moved Modern UI Apps or SkyDrive Sync to your SD Card? Used Sym-Links from C:\Folder to your SD Card? I haven't been able to replicate the error using my Sandisk Ultra 64GB Class 10...

no my SD card only has music files on it, not added to "Libraries" either
 
Yes I can replicate this every time by resuming from hibernation with the SD card installed, CPU does not go up with SD card out on resume from hibernation. Thanks for the insight! So... people with SD cards, force a hibernation "shutdown /h" and resume (power button) with SD card in, do you get high CPU utilization? (22% on Balanced, 30+% on High Performance)

I have a few SD cards of different types and will try them when I get home to see if its a compatibility issue with this specific type of card.

Yes, I've had a very similar issue. With an SD card in the slot, shut down the computer, when off for > 15 sec, power on, and invariably the CPU utilization will be 25 or 33% depending on power plan in effect. This does not happen with a "restart" of the computer, only with full shutdown and power up. I'm guessing that resuming from hibernation is essentially the same thing.

One "trick" I've noticed is putting the SP2 into sleep mode and then waking from sleep will result in CPU use settling to the normal near 0% level. I'm quite convinced this has nothing to do with the mSD card file system or contents, but rather a firmware bug whereby the mere presence of the mSD card results in an "interrupt storm" and high activity in kernel processes.

I'm in the habit of checking taskmgr after I log in, and if CPU activity is 25% or more, hit the sleep button, and immediately awaken. Taking care with power options and what I'm running on the SP2, battery life is good, averaging ~9 hours per powercfg.
 
See the top issue here: Surface Pro 2 Issue List - Microsoft Community, lots of reports of the same issue, think it's been know about for a while.

Likely compatibility with your SD card, not a dud surface that a replacement would fix so maybe try a different card while wasting for MS to get into gear.

Edit: Spellchecker changed 'waiting' to 'wasting', but sounds appropriate given the circumstances so think I'll keep it as it is :)

I've read through the above thread, actually posted in it a while back. The consensus seems to be this is very widespread and easily duplicated issue, it may have to do with the type of SD card even though many, including myself, have SD cards marked compatible with SP2. An interesting question posted that I am curious about myself is whether or not this issue carries over to the newer i5 4300u hardware revision? I know its a long shot it doesn't but I'm still very curious.
 
I've read through the above thread, actually posted in it a while back. The consensus seems to be this is very widespread and easily duplicated issue, it may have to do with the type of SD card even though many, including myself, have SD cards marked compatible with SP2. An interesting question posted that I am curious about myself is whether or not this issue carries over to the newer i5 4300u hardware revision? I know its a long shot it doesn't but I'm still very curious.

mSD card induced CPU revving does occur on my 4300u SP2. Curious if anyone else has success with the "sleep on/off" trick. Works reliably on my SP2--quiets down the CPU utilization to normal level.
 
Occurred to me that I could test with the microSD card in my Android phone. Mounted a 32 GB Patriot SDHC with a 10 rating and formatted FAT32, so far couldn't replicate the issue. May be old school but this reminded me a lot of an IRQ conflict, checked and didn't see any with the SD card mounted,. Guess I could play with it a bit more to see if I can make my SP2 behave like many using SD cards have reported.


Okay was able to peg the CPU to 24% with the SD card mounted, to quiet the CPU down I had to un-mount the SD card and do a restart. I wonder if Windows is running drive optimizations?
 
Last edited:
Occurred to me that I could test with the microSD card in my Android phone. Mounted a 32 GB Patriot SDHC with a 10 rating and formatted FAT32, so far couldn't replicate the issue. May be old school but this reminded me a lot of an IRQ conflict, checked and didn't see any with the SD card mounted,. Guess I could play with it a bit more to see if I can make my SP2 behave like many using SD cards have reported.

Wondering exactly how you tested the mSD card issue on your SP2? I've found the CPU revving (>=25% CPU utilization) when starting the SP2 from full power off, but not otherwise, e.g., waking from sleep, or doing a "restart".

Not sure that the characteristics of the card or its contents have any effect on the problem at all.
 
I edited my post, I was eventually able to get the SP2's CPU to peg at 24% useage with a SD card mounted. Looks like currently the SP2 doesn't play nice with SD cards.

Or possibly a Windows 8.1 issue. Now I'm curious as to how my X1CT will handle adding a SD card.
 
Last edited:
I edited my post, I was eventually able to get the SP2's CPU to peg at 24% useage with a SD card mounted. Looks like currently the SP2 doesn't play nice with SD cards.

Or possibly a Windows 8.1 issue. Now I'm curious as to how my X1CT will handle adding a SD card.

it would be a win 8.1 on SP2 issue at most but its odd not everyone can replicate this
 
One "trick" I've noticed is putting the SP2 into sleep mode and then waking from sleep will result in CPU use settling to the normal near 0% level. I'm quite convinced this has nothing to do with the mSD card file system or contents, but rather a firmware bug whereby the mere presence of the mSD card results in an "interrupt storm" and high activity in kernel processes.

I'm sure you're right. I've now reproduced this, and nether sysinternals/procmon or the windows task manager/resource monitor show any file activity so we can probably discount any application/virus scanner/indexer being the fault here.

All it shows is an interrupt handler (ISR) and the associated deferred procedure call (DPC) Looks like the hard power up sequence is dodgy and some sticky error or interrupt has been raised and is never cleared down.

FWIW, I used to write SD card drivers for a mobile phone company and have encountered similar issues many times. If MS would like to pass me their code and reference platform I'll gladly take a look and have it fixed over the weekend.
 
Yes, I've had a very similar issue. With an SD card in the slot, shut down the computer, when off for > 15 sec, power on, and invariably the CPU utilization will be 25 or 33% depending on power plan in effect. This does not happen with a "restart" of the computer, only with full shutdown and power up. I'm guessing that resuming from hibernation is essentially the same thing.

One "trick" I've noticed is putting the SP2 into sleep mode and then waking from sleep will result in CPU use settling to the normal near 0% level. I'm quite convinced this has nothing to do with the mSD card file system or contents, but rather a firmware bug whereby the mere presence of the mSD card results in an "interrupt storm" and high activity in kernel processes.

I'm in the habit of checking taskmgr after I log in, and if CPU activity is 25% or more, hit the sleep button, and immediately awaken. Taking care with power options and what I'm running on the SP2, battery life is good, averaging ~9 hours per powercfg.

Exactly the same happening to me! I wasn't even aware of that issue, but when I read this post, I checked the task manager and found my processor working at 20-30%. I removed the Sd card, rebooted and then it worked at 0-1%. If I add the Sd card again, nothing weird happens, but if I completely power off the computer for 15s and then start it up with the SD inserted, it idles again at 20-30%.
There are two processes making the processor work: one called System (it uses 15% of the processor) and another one called System Interruptions (using about 8% of the processor)

Another hell of an issue to add to my issues list... though this one seems to have a simple solution: just don’t use the Sd card on a 1000€ device...
 
Last edited:
Exactly the same happening to me! I wasn't even aware of that issue, but when I read this post, I checked the task manager and found my processor working at 20-30%. I removed the Sd card, rebooted and then it worked at 0-1%. If I add the Sd card again, nothing weird happens, but if I completely power off the computer for 15s and then start it up with the SD inserted, it idles again at 20-30%.
There are two processes making the processor work: one called System (it uses 15% of the processor) and another one called System Interruptions (using about 8% of the processor)

Another hell of an issue to add to my issues list... though this one seems to have a simple solution: just don’t use the Sd card on a 1000€ device...


This is terrible, its been an issue since release and nothing's been done about it yet. Furthermore, actually using the now fixed hibernation after sleep for 60 minutes feature is still not recommended since it causes this issue. Also, I tested this with 3 more microSD cards I've collected over the years (various classes/sizes) and the same thing happens on all of them. Again I ask, is this universal? Do I have a lemon? Or can every SP2 replicate this from a cold-boot (powered off or hibernation) It makes me mad to consider the possibility that others have uncompromised hardware where this doesn't happen and my brand new $1500 SP2 256 (w/ accessories) has a defect.
 
I've never seen this on any of the 3 Surface Pro 2's I support (Mine, Wife's and Son's) All are original 256/8 models. Mine idles between 0-10% CPU with 42% Memory Utilization.

Please test this from cold boot (wait 15 sec after shutting down) with an sd card inserted. Also please let the community know what SD card you are using, thanks
 
Back
Top