graphite benchmark

Did a little benchmark using graphite on the whetstone source. As you can see graphite makes a huge difference here. Even on a single-core machine.
The source of the whetstone benchmark can be downloaded here: http://www.netlib.org/benchmark/whetstone.c
I have not modified the source in any way.

Machine specs:
HP Pavilion a1220n
CPU: 2.93Ghz Pentium-4 516
RAM: 3GB DDR2
Kernel: Linux jennifur 3.4.2-ck2 #1 PREEMPT Tue Jun 12 13:37:24 EDT 2012 x86_64 Intel(R) Pentium(R) 4 CPU 2.93GHz GenuineIntel GNU/Linux

I have compiled whetstone-graphite using:
gcc whetstone.c -o whetstone-graphite -lm -march=native -floop-interchange -floop-strip-mine -floop-block -fgraphite-identity -O2


root@jennifur:/home/jennifur# time ./whetstone 100000

Loops: 100000, Iterations: 1, Duration: 6 sec.
C Converted Double Precision Whetstones: 1666.7 MIPS

real    0m6.066s
user    0m6.065s
sys     0m0.000s
root@jennifur:/home/jennifur# time ./whetstone 100000

Loops: 100000, Iterations: 1, Duration: 6 sec.
C Converted Double Precision Whetstones: 1666.7 MIPS

real    0m6.042s
user    0m6.040s
sys     0m0.000s
root@jennifur:/home/jennifur# time ./whetstone 100000

Loops: 100000, Iterations: 1, Duration: 6 sec.
C Converted Double Precision Whetstones: 1666.7 MIPS

real    0m6.073s
user    0m6.071s
sys     0m0.000s
root@jennifur:/home/jennifur# time ./whetstone-graphite 100000

Loops: 100000, Iterations: 1, Duration: 2 sec.
C Converted Double Precision Whetstones: 5000.0 MIPS

real    0m2.285s
user    0m2.266s
sys     0m0.000s
root@jennifur:/home/jennifur# time ./whetstone-graphite 100000

Loops: 100000, Iterations: 1, Duration: 2 sec.
C Converted Double Precision Whetstones: 5000.0 MIPS

real    0m2.265s
user    0m2.264s
sys     0m0.000s
root@jennifur:/home/jennifur# time ./whetstone-graphite 100000

Loops: 100000, Iterations: 1, Duration: 2 sec.
C Converted Double Precision Whetstones: 5000.0 MIPS

real    0m2.265s
user    0m2.264s
sys     0m0.000s