DAW stands for Digital-Audio-Workstation.
I'd been using SETI@Home full time in back ground of DAW applications such as Steinberg Nuendo (it does hard disk recording and realtime DSP calculations.
I've never had any trouble with SETI, it runs quite well in the back ground and doesn't affect fore ground applications.
but since i switched to BOINC (Einstein@Home), Nuendo runs slowly and CPU usage meter of the Nuendo shows 50%(?) more activity than normal. it also affects the realtime DSP power, you can only use 50% less DSP when BOINC is running in behind even all the processes are set to lowest priolity. It also often causes audible gliches in sound output. All these phenomenon makes BOINC useless for me.
This wasn't happening with the classic SETI at all, it was always running background for the last 5? years or so, on various platforms i have used.
Can anyone tell me what has changed and why?
Are there any chance that BOINC will be as audio-application friendly as the classic SETI? Otherwise i have to remove it from my system, that is not what i want to do.
Thanks in advance, any help appreciated.
Takashi Watanabe / Sound Engineer
Copyright © 2024 Einstein@Home. All rights reserved.
BOINC conflicts with DAW applications
)
I don't know of DAW applications, but did you have set "Do work while computer is in use" in your preferences? If yes, you might want to set it to "no". E@Hseems to be more CPU- (actually FPU-) bound than SETI, so it's usually not advisable to run it beside other applications, especially if they are time-critical. Windows isn't a real-time operating system, so every real-time program needs to tweak it somehow to get it to work as expected. This might conflict with E@H or the idle detction mechanism of BOINC.
BM
BM
I've been experiencing
)
I've been experiencing similar difficulties with E@H and iTunes playing music - there an awful lot of stuttering, sometimes resulting in iTunes giving up and waiting for me to hit the play button again.
The problems seem to occur at regular intervals (about 15 mins), suggesting that maybe there's a cyclic problem. Bruce Allen speculated that it might be due to the large checkpoint files that E@H writes. I've seen the problem on different machines, but I haven't seen it when LHC@Home units have been running.
'Do work while computer is in use?' set to 'no' might be a partial fix, but if I'm using the computer to play music while I eat then I don't want to have to tap the keyboard every few minutes to fool BOINC. I may have to detach my music library machine from E@H if I can't find a better solution
Mike W