scala vs. java in calculation on c2d

lauri@derp-server:~/java/recursiontest$ time java Assigment
run 8004750 times
2000,3750,4250.
1810228928

real	0m0.549s
user	0m0.531s
sys	0m0.034s
lauri@derp-server:~/java/recursiontest$ time java Assigment
run 8004750 times
2000,3750,4250.
1810228928

real	0m0.534s
user	0m0.532s
sys	0m0.024s
lauri@derp-server:~/java/recursiontest$ time java Assigment
run 8004750 times
2000,3750,4250.
1810228928

real	0m0.534s
user	0m0.525s
sys	0m0.024s
lauri@derp-server:~/java/recursiontest$ time scala ~/scala/recursion.scala
run 8004750 times
List(2000, 3750, 4250)
1810228928

real	0m3.058s
user	0m2.022s
sys	0m0.182s
lauri@derp-server:~/java/recursiontest$ time scala ~/scala/recursion.scala
run 8004750 times
List(2000, 3750, 4250)
1810228928

real	0m3.181s
user	0m2.073s
sys	0m0.163s
lauri@derp-server:~/java/recursiontest$ time scala ~/scala/recursion.scala
run 8004750 times
List(2000, 3750, 4250)
1810228928

real	0m2.967s
user	0m2.026s
sys	0m0.193s
lauri@derp-server:~/java/recursiontest$ javac Assigment.java
lauri@derp-server:~/java/recursiontest$ time java Assigment
run 800047500 times
20000,37500,42500.
2047696384

real	0m32.709s
user	0m32.612s
sys	0m0.114s
lauri@derp-server:~/java/recursiontest$ time scala ~/scala/recursion.scala
run 800047500 times
List(20000, 37500, 42500)
2047696384

real	1m26.455s
user	1m25.088s
sys	0m0.637s
lauri@derp-server:~/java/recursiontest$ time scala ~/scala/recursion.scala
run 800047500 times
List(20000, 37500, 42500)
2047696384

real	1m26.108s
user	1m24.956s
sys	0m0.621s


notes:

	*java used for-loop
	*scala used recursive function

core2duo usage:

	*java would saturate single core for the calculating

	*scala would use both cores but both only up to ~50% value (except immediate start), total would be just under 110% all the time, maybe more cores would make it more effective?