Request installation option to turn-off hardware acceleration completely

mirillis

Administrator
Staff member
It only consumes CPU when in remotely session. And killing it breaks the session.
During start, Remotly.exe gives a signal to remotly_boot.exe to close so a hand-over can be done. For some, yet unknown reason, this does not work in this case.
 

hhspiny

Member
Follow-up on remotly_boot.exe. with the 1.34 pre-release, which seems to behave the same as 1.33.1
Note, this is in windows 11 (not VM), installed with admin account, start with windows. but do NOT start remotely client in normal user account (i.e. normal user account still uses the admin instance)

when admin account is logged in, remotly_boot.exe closes itself, not in the process list. (processes: remotly_x86/64.bin, remotly_svc.exe, remotly_logon.exe, remotly_launcher_x64.exe, remotly_boot_svc.exe, remotlyLauncher.exe, remotly.exe)

when normal user account is logged in, remotly_boot.exe remains in process list. (processes: remotlyBootHelper_x64.exe remotly_boot.exe, remotly_boot_svc.exe, remotly_logon.exe, remotly_svc.exe)

runs without problem though. the only problem that noticed as before, after long network interruption, host won't log back in. and sometimes get blank-screen at connection which not sure if is related.
 

mirillis

Administrator
Staff member
Hi hhspiny,

Thank you for your prompt reply!

"runs without problem though. the only problem that noticed as before, after long network interruption, host won't log back in. and sometimes get blank-screen at connection which not sure if is related."

If host won't log back in then it's a critical error :( I've also had this problem once with fully logged in booter. Seems not to happen for unlogged booter in QuickHost (QuickHost works only with Connection ID codes). I also got blank-screen once or twice, but reconnecting seems to solve the problem so it's not a critical.

Seems QuickHost is more reliable for administrative purposes (never had any problems connecting to QuickHost, even with serious network interruptions).
It also blocks all connections outside organization (even knowing Connection ID of a specific host).
 

mirillis

Administrator
Staff member
The log back in problem might be related to this:

Final v1.34.0 update should finally resolve this.
 

mirillis

Administrator
Staff member
Not so good on our side with GPU-P on Nvidia with latest drivers. Hardware acceleration dissapears on RTX 5000 series. We are working on this right now. Nvidia does not make any regression tests. What's interasting is that these problems are happening now when they are replacing humans with AI for various tasks. Some things that used to work flawlessly for the past 10 years are not working anymore.

Some guys from our team are currenly in Amsterdam and take feedback from current and new users. They are showcasing high performance simultaneous connection to two VMs with GPU-P on one host. And a bunch of other things.
BTW. We will be also present on NAB in Las Vegas April 2026. This time with support for Apple product too.

IMG-20250912-WA0012.jpg


To make things clear. It's not me in the picture...
 
Top