Windows' CPU Timer Resolution at "lowest" after VPN connected?
- February 25, 2017
- 1 reply
- 5466 views
Hi Everyone,
I'm here to seek your kind help in understanding what's going on between "VPN connections" and "Windows' CPU Timer Resolution", and especially when it's connecting via FortiClient (latest, 5.4.2.0860).
I recently noticed that whenever I'm connected to VPN, my notebook battery seems to be draining much faster than when I'm not. Then after some googling I've come up with a possible cause related to "CPU Timer Resolution" of Windows (I'm on Win10 - 1607 - 14393.693)
ClockRes (a utility to display the current Timer Resolution)
https://technet.microsoft...nternals/bb897568.aspx
When VPN is connected via FortiClient in my case, CPU Timer Resolution would *persistently* go to ~5ms (fine), as oppose to the disconnected state at ~15ms (normal idle) (see attached screenshot).
To my limited knowledge, unnecessarily fine/short timer resolution would drain battery, just like what Chrome did some versions ago:
https://arstechnica.com/g...raining-bug-in-chrome/
https://bugs.chromium.org...ssues/detail?id=153139
I wonder if a Timer Resolution this fine/short (~5ms) is necessary for any generic VPN connection, or is it specifically set by FortiClient? Or maybe it's just a bug of FortiClient or just of Windows or the Network Card driver?
I've just tested with Chrome (latest) that they have fixed their bug and now it won't shorten the timer resolution unnecessarily and now battery doesn't drain as quickly as before.
This really bugs me since my work requires me connected to VPN most of the time. If a fine timer is necessary for any VPN connection then I might need to search for another solution on the OS level. But if it's caused by FortiClient's setting then maybe I could report a bug or consult the dev team for some special workaround?
Thanks very much for your insight.
Regards,
Godric