There was a thread recently pointing to a discussion on GearSluts where a Reaper user was shocked that Sonar was running a VST test with much better results than Reaper. The Reaper user went through several interations of changes to elements of the test (even to the point of using a totally different VST) until Reaper was coming out on top. At that point the discussion ended because the Reaper user had gotten the answer they were looking for. So, perhaps you too can follow his lead and find a test that gives you the answer you are looking for.
 
http://www.gearslutz.com/board/music-computers/887142-dawbench-sonar-x3c-vs-reaper-4-57-a.html

I've tried various permutations on multiple individual channels and busses to multiple instances on one channel.There is no denying that some Nebula libraries like plate reverbs devour CPU no other plugin I have sucks the life out of the processor like it but I prefer Nebula over almost everything else.It would be great to be able to use it in Sonar but either I'll track in X3 and mix down on Reaper or I'll switch over to Reaper completely which I'm reluctant to do as I'll have to develop a new workflow.A new computer with a lot more grunt would also be a solution but like many people after Christmas I'm skint so that's not on the menu for a while.

bentleyousley
There was a thread recently pointing to a discussion on GearSluts where a Reaper user was shocked that Sonar was running a VST test with much better results than Reaper. The Reaper user went through several interations of changes to elements of the test (even to the point of using a totally different VST) until Reaper was coming out on top. At that point the discussion ended because the Reaper user had gotten the answer they were looking for.
 
http://www.gearslutz.com/board/music-computers/887142-dawbench-sonar-x3c-vs-reaper-4-57-a.html

 
This is something I find useful that is quick and dirty, still not at all like a proper test, but often gives an overview:
http://technet.microsoft.com/en-gb/sysinternals/bb896653.aspx
 
That's not to say a space station is a Ferrari.

Have you considered freezing tracks after you record them so only a few VI instances need to be using the CPU at any one time?

I've tried freezing and obviously it works but I feel it slows the workflow down as i like to tweak a lot.It feels like I'm using a 4 track again when I use freeze:)

What seems really odd to me isn't so much that Reaper may be a bit more efficient overall than SONAR - I can buy that - but rather that it's 16 vs. 6 instances of Nebula.  That's a 266% difference!  I just can't understand how one DAW could be anywhere near 2.66X more efficient than another, all (other) things being equal.
 

Steve that's exactly why I posted this thread.As soon as I tried Nebula in Reaper I was gobsmacked and my 1st reaction was what am I doing wrong in Sonar.In fact what prompted me to try Reaper was that Henry Olonga who produces some well respected Nebula libraries and who is also a Sonar user wrote on his website blog about how he tried Reaper and got much better results with Reaper than with Sonar.Now it may be that Sonar just doesn't get along with Nebula libraries but seeing is believing and Reaper slays Sonar when using this plugin at least on my system( and Henry Olonga's)

stevec
What seems really odd to me isn't so much that Reaper may be a bit more efficient overall than SONAR - I can buy that - but rather that it's 16 vs. 6 instances of Nebula.  That's a 266% difference!  I just can't understand how one DAW could be anywhere near 2.66X more efficient than another, all (other) things being equal.

Because it's different software and a totally different environment, even if you have the same computer/drivers etc.
You just cannot compare DAW's this way. One is chalk, the other is a cow, the other is Pamela Anderson.... that's just the way it is.
 
I guess it's a little like probability theory, there's some stuff here that's seems like it couldn't possibly be true but it is. If you used just your eyes on some of this stuff you would be wrong, consistently, in all sorts of scenarios.

I understand that each DAW is unique in the way it processes plugins, audio, etc.   But a 2.66X difference seems to indicate something beyond just that.  I mean we're talking (hypothetically) 38% in one DAW vs. 100% in another (assuming one can even run at 100%).  Or maybe SONAR and Nebula simply don't get along, although it doesn't ring any bells.
 
Pict,
What is the CPU usage/percentage of one instance in both DAWs, across all cores that are showing activity?   How about with six instances? 
 

I'd be interested in how other plugins would perform or if it is only Nebula.