Generally, you would expect execution times to be different for dual core Vs. single core in a multi-threaded scenario (either a multi-threaded application or multiple processes). Shorter execution times are due to true execution concurrency (or
parallelism in Intel-speak).
The first question might be, how do you know that the process isn't just taking more time to execute and that the timer function is working correctly?
If NI is using performance timers there could be an issue. There are reports of QueryPerformanceCounter being goofed on multicore AMD's. Also RDTSC Pentium instruction has multicore issues too apparently. And the Intel SpeedStep technology, which clocks down the CPU when it is idle, could goof timing calculations depending upon the implementation.
This is interesting especially since NI is claiming milticore prowess with CVI 8.5.
Menchar