Hardly anyone has mentioned lower CPU usage!

I have been waiting since I installed and started using X2 to see any posting about the lower CPU usage. I have noticed on older projects from 8.5 and Expanded, the CPU is significantly reduced! The other threads in my quad seem to be working more consistently as well. Has anyone else noticed this? Like I said, I haven't seen too much said about this. I know that the engine has been tweaked a bit but this is showing a significant difference. Anyone else?

I honestly can't say that I've noticed Sidroe. When I first installed X1 a couple of months ago I noticed that my first CPU thread got the most hammer and then perhaps the second thread. The other six threads were a bit erratic. I found a post here about "core parking" and I turned this Windows feature off. Since then, I still find that my first CPU thread gets more use, but the remaining seven threads are a lot more equally used.

In X2 I have not noticed any improvement on this.

cheers

andy

I agree that the first core is still taking the biggest hit. It looks like the hit is not near as big as in X1 and the other cores are more active.

Sidroe, no, you'll want to *disable* core parking (if it's enabled). It's probably already disabled. Here's one of the forum threads on it: http://forum.cakewalk.com/tm.aspx?m=1861804

I have a torture-test of an X1 project that I opened in X2 and did not notice any CPU utilization changes either in raw numbers or distribution across the cores. Other projects might vary, of course, but that's my report.

unganged is best for quad-cores and for multi-threaded applications. ganged is best for single-threads that use a lot of memory. to my knowledge, Sonar does a bit of both. i don't think i'll will matter much. i would probably stay with unganged as most apps are multi-threaded these days.

but, the difference is very small either way. and it's about memory bandwidth which has very little to do with audio engine performance. the biggest thing for Sonar is CPU (the floating-point benchmarks).

even for gaming where you have a lot of texture swapping cases where memory bandwidth matters, the benchmarks i've seen where like 0.1 percent different. there are many other factors and settings to worry about that will have a more dramatic effect on performance.

Kenneth


In X1 on the same template I've noticed that the first core got hammered  to 25-30% more than the others, now it's more spread out and the first core doesn't seem to take such a high load.

I notice my first core spiking too, to 2 to 3 times as high as the other cores. If I un-mute the z3ta+2 track, all the cores double in height. This happens in X2 as well...

Just for fun, take a look at your configuration file in the preferences. See what your thread scheduling model is set to between 8.5 and X2. They should be the same since X2 imports the setting from 8.5 when installed but perhaps they are different. I have an i7 system and I have it set to "2". There is definitely different results between the three settings. 

Here is a rundown on the setting:

ThreadSchedulingModel=<0 - 2>  Type Integer  Default 1  This variable goes in the [Wave] section and controls the interaction of the main audio thread and worker threads on multiprocessor systems when the Use Multiprocessing Engine option is enabled. Depending on the system, a particular model may result in less glitching and better overall performance. The values are as follows: 


0 = Same as previous versions of SONAR. 
1 = (default) Better thread balance. Model is more efficient and can provide cycles for other tasks. 
2 = Additional worker thread is created. This may result in improvement with Quad processor systems or higher. Not recommended for Dual processor systems. 

Lower CPU usage & better core spread here even on XP!
HeyDan

Jalcide


unganged is best for quad-cores and for multi-threaded applications. ganged is best for single-threads that use a lot of memory. to my knowledge, Sonar does a bit of both. i don't think i'll will matter much. i would probably stay with unganged as most apps are multi-threaded these days.

but, the difference is very small either way. and it's about memory bandwidth which has very little to do with audio engine performance. the biggest thing for Sonar is CPU (the floating-point benchmarks).

even for gaming where you have a lot of texture swapping cases where memory bandwidth matters, the benchmarks i've seen where like 0.1 percent different. there are many other factors and settings to worry about that will have a more dramatic effect on performance.

Wow capow thanks Jalcide. I'm gonna sell my shovel and buy a JCB when I get rich.