I had never heard of TBB before reading this so I went and found an Intel description. It really sounds like a nice product for taming a non-realtime OS where you want to write a large multi-threaded do everything process that needs to priorities things. It does it’s own internal scheduling.
No offense intended, but this seems like nothing you would want to use with QNX, unless you don’t know what you are doing.
The description suggested that the library would help you avoid having too much competition for the available cores, and it would let you prioritize threads into varying categories of importance. The only way this would be of help in QNX would be if you run a single process with lots of cpu bound threads, all at the same priority. And if you are doing this on QNX, (apologies) you don’t know what you are doing.
Sorry I’m not more helpful.