Recent Posts

Pages: 1 ... 5 6 [7] 8 9 10
61
General / Asus still doing boring thing on CPU for touchpad..
« Last post by BenYeeHua on September 10, 2016, 08:06:29 PM »
After they modified the Elan driver to force running highest frequency, then removed it after a few version, they did it again on their SmartGesture...

I can said it is the Reviewer causing it, as they said it is not smooth as Mac, so they force it at highest frequency while scrolling, but they write the wrong code and it is doing even when it is not scrolling. :P

Quote
I ask why, for goodness sakes, why is this there? And the guy explains in a roundabout fashion that support has been contacted, and reviews have mentioned, that the asus scrolling is not as "smooth" as the scroll in for example a Mac.
http://forum.notebookreview.com/threads/enabling-3-finger-tap-middle-click-and-other-functions-on-elan-touchpads-zenbooks-etc.684036/page-3#post-10326662

I think the not "smooth" issues has been fixed by Intel ParkControl, but I still thinking the issues is, the reviewer is using Chrome for testing, but Chrome don't enable smooth scrolling yet... ::)
https://bitsum.com/product-update/intels-speed-shift-validates-parkcontrol-and-process-lasso/

And also, the polling rate is too low, as far as I get in the ETDDeviceInformation.exe, it is 78(hz) report rate, which is 12ms per report, while the Screen is having 60hz(16.67ms) refresh rate.
So, in order to scrolling, you need to have 2 report telling the driver you are scrolling, which is 12+12ms=24ms, and based on Firefox and Chrome Canary's smooth scrolling, this smooth scrolling has added at least 2 frame late for the scrolling effect.
---
So I just checked the Windows Precision Touchpad, they are asking for at least 100HZ input for multiple contacts(multiple finger), which is 10ms.
Quote
User-perceived latency significantly diminishes the experience of a Windows Precision Touchpad and therefore all aspects of the system from end-to-end shall meet or exceed specified latency goals. Providing a minimal input report rate of 125Hz for single contacts and 100Hz for multiple contacts ensures that with the correct scan frequencies, contact down and update latencies of 25ms and 15ms respectively can be achieved.
https://msdn.microsoft.com/en-us/library/windows/hardware/dn467309(v=vs.85).aspx

And as you can see, it is I2C, as I just download a few driver and found some of them are I2C driver inside, they(ELAN and ASUS) did just handle the Precision Touchpad with their own solution, but not handle it to Windows as it lack of configuration for the user to change. :P

But then they just doing this boring thing on the driver, I hope they don't do this for CPU at least Skylake Gen. ::)

As I don't has money to buy a new laptop and do this research, so I guess I will just stop here, and waiting other user to tell the story. 8)
62
General / A few change on Windows 10 1607
« Last post by BenYeeHua on September 10, 2016, 07:38:07 PM »
Just some change that I observed.

The Adobe Flash Player will calling the Multimedia Scheduler Driver to set the Timer Resolution as 0.5ms, which causing your CPU running 95% non-stop, and wasting your power. :P

While the Edge will compress the Memory and showing as Paged Pool in Windows Task Manager, and freeze itself sometimes. :o

And also, WDDM 2.1 is included, but most graphic card driver not catching up yet. ::)
https://en.wikipedia.org/wiki/Windows_Display_Driver_Model#WDDM_2.1

That's all. ;D
63
General / Useful Memory Compression that might make you hard to debug memory issues
« Last post by BenYeeHua on September 10, 2016, 07:32:26 PM »
Just saying sometimes that happen on my laptop, so as I went to upgrade to Windows 10 1607, I found a new touchpad driver and update to it.
And well, it(ETDCtrl.exe) was memory leak silently on the background, and lucky it is empty data, while unlucky Windows 10 choose to compress it...

Then a week ago, I just saw the strange commit size is increased while the RAM usage never increased, the PageFile usage is also very low, and it is showing it has compressed 1-3GB of data, which is a lot! :o
The leak is also not showing in the ETDCtrl.exe private or working set, it is not showing as Paged pool as it is not leak by driver, I think it might be not showing because it is compressed... ???
So the only clue that I has, is I updated this driver while I update Windows 10 1607. :P

So ya, I think MS should provide a function to disable Memory Compression via BCDedit, for debug, not just silently compressed leak empty data, and not letting the user know.  ;)
64
General / Re: SuperFetch, useful or useless for Individual users?
« Last post by BenYeeHua on September 10, 2016, 07:17:24 PM »
And I think it(readyboot) might causing issues in Windows 10, sometimes I has a fast boot/shutdown computer, sometimes it is just waiting the hard disk for 10mins during normal boot/fast boot shutdown(means hibernate the kernel+driver only).. :P

All of this issues gone if I force shutdown the computer once, or going into safe mode to let itself clear some hidden cache, and I think the cache is this readyboot.
---
I forgot to update this post with new Win10 and AU (1607 build), as Chris mentioned early, MS now has superfetch key gone and the readyboot folder in windows is empty now, so only prefetcher reg key is 3
Strange, the readyboot folder is not empty for me, and it is having the modified record 20mins ago, which is when I reboot my computer. :o
Maybe the settings left over from Windows 8 to Windows 10?

And ya, for the superfetch, it is a bit stupid now, after you just bootup your computer, you run your heavy Chrome browser, it will bite your harddisk and causing slow boot of Chrome.
I observed the superfetch just kick in after finish booting, I think it should having the low I/O priority, so I wonder why it will just slow down the computer.

Maybe Chrome is too heavy for cold boot then, as they need to set difference prefetch record for difference processes.
Quote
/prefetch:# arguments to use when launching various process types. It has been observed that when file reads are consistent for 3 process launches with the same /prefetch:# argument, the Windows prefetcher starts issuing reads in batch at process launch. Because reads depend on the process type, the prefetcher wouldn't be able to observe consistent reads if no /prefetch:# arguments were used. Note that the browser process has no /prefetch:# argument; as such all other processes must have one in order to avoid polluting its profile. Note: # must always be in [1, 8]; otherwise it is ignored by the Windows prefetcher. ↪
http://peter.sh/experiments/chromium-command-line-switches/
65
General / Re: SuperFetch, useful or useless for Individual users?
« Last post by edkiefer on September 09, 2016, 03:58:11 PM »
I forgot to update this post with new Win10 and AU (1607 build), as Chris mentioned early, MS now has superfetch key gone and the readyboot folder in windows is empty now, so only prefetcher reg key is 3
66
Tips and Tweaks / Re: Trying to identify Acronis notifiyer
« Last post by BenYeeHua on September 05, 2016, 09:26:38 PM »
If the one that posted by Hotrod is not working, try this LogWindowAtPoint, it should works, unless the windows pop-up and auto close it.
hxxp://www.uwe-sieber.de/misc_tools_e.html

PS: Remember run it with Admin right :)
67
Tips and Tweaks / Re: Trying to identify Acronis notifiyer
« Last post by Hotrod on September 05, 2016, 11:44:30 AM »
I just googled "Acronis upgrade notifier" and got this :) https://kb.acronis.com/content/57373

Hope it has your answer :)
68
Tips and Tweaks / Trying to identify Acronis notifiyer
« Last post by danielson on September 05, 2016, 09:54:39 AM »
Acronis True Image is nagging me to upgrade and i don't want to.

How can i get Process Lasso to stop it?
It is hard to identify the process that is behind this.
69
Tips and Tweaks / Re: Skype on Win 10 chews at cpu
« Last post by danielson on September 05, 2016, 09:51:51 AM »
Set to idle = yup, that is what i eventually did.  8)

Soorry did not respond before, thought i had automatic notification in settings.
70
Process Lasso / Re: weird update msg
« Last post by bertie97 on September 01, 2016, 05:57:57 AM »
Thanks for the replies.

I do have the VC++ packages installed & up to date AFAIK.  I sometimes need these for the *very rare* gaming sessions ;)
I am using Comodo CIS.   
Normally whenever the PL updater comes up I run a CCleaner instance & turn off the CIS sandbox, before hitting 'update' to avoid any mess/interference.
In this case I had shut everything down, & cleaned up ready to power down the machine, & the updater came up just before I actually hit shutdown.
Technically there shouldn't have been anything much running, certainly nothing that wasn't running on the last update.

As far as I can tell PL is working OK, I've been running it for a few hours now & it seems to be fine.  So a bit weird but relatively uneventful... :)
Pages: 1 ... 5 6 [7] 8 9 10