Thank you for clearing that up.

For single core hyperthreading systems kernelopts(numcpus) is 1.  Maple sets the number of gcmaxthreads to numcpus or 4 whichever is lower (in my case 1).  I decided to see what effect the timings would have if any when I changed the number of threads, I expected no real change but in fact got slightly better timing.

       Error, unable to execute seq

I recieved an error using shown seq.  So instead I used  data := [seq(i, i = 1 .. 10000000)] to test the timings out on my machine.  It is a P4 3Ghz single core hyperthreading 2Gb RAM WindowsXP 32 bit system.  Since hyperthreading CPU's are treated as 1 cpu, gcmaxthreads by default would be set to 1, however increasing gcmaxthreads to 2 gave on average a .5 second better timing.

gcmaxthreads=1 timing was around 8.750
gcmaxthreads=2 timing was around 8.200. 

The decrease in timing was very slight when gcmaxthreads=4 and sometimes more than when threads were 2 depending on if the computer was thinking in the background but gcmaxthreads=2 could not better gcmaxthreads=4 for the best timing. 

It is possible that hyperthreading systems can gain slightly better timings by increasing your threads above the default value.