Benchmarking the Raspberry-Pi

This is a crude and simple PI test against the Raspberry-Pi and a Dell Pentuim-4 (3.0Ghz)

The Raspberry-Pi
Starting PI…
x= 0.38631 y= 0.89070 low= 939239 j=1200001
Pi = 3.130797 ztot= 801773.75 itot= 1200000

real 0m1.900s
user 0m1.720s
sys 0m0.000s

Pentium-4
Starting PI…
x= 0.38631 y= 0.89070 low= 939239 j=1200001
Pi = 3.130797 ztot= 801773.75 itot= 1200000

real 0m0.099s
user 0m0.096s
sys 0m0.000s

This somewhere between 17x and 19x speed difference, but compairing a ARM RISC ALU with a Pentium’s CISC FPU is not a fair comparison either. In other testing I was doing, I only got a 5x difference in performance, roughtly the difference between the Raspberry-Pi’s 700Mhz and the Pentium’s 3.0Ghz clock frequency.

As I got this benchmark from an old site, I was amused to note that this benchmark only makes the Raspberry-Pi about 56x times FASTER than a MicroVAX-II, a system I cut my teeth on in my programming life.

Life’s relative, maybe I should write that down.

Note; the GPU was NOT part of the testing.

UPDATE: I managed to get this bencmark to run on a Arduino Uno (16Mhz) and the Pi is 126 times faster, So much for using an 8 bit processor as a number cruncher 🙂

UPDATE 2:

The new Raspbian Distribution, (after a recompile) produced this benchmark.

Starting PI…
x= 0.38631 y= 0.89070 low= 939239 j=1200001
Pi = 3.130797 ztot= 801773.75 itot= 1200000

real 0m0.539s
user 0m0.520s

sys 0m0.010s

or between 3.52 and 3.30 TIMES faster that the original Debian Squeeze distro. That makes it only about 5.44 times slower than the Pentium 4 at 3.0Ghz which I was detecting before with non numeric benchmarking. An interesting side note, due to the ‘Hard Float’ the timings incured 1/100 second of SYS time on the benchmark.