RC5 Long Benchmarks

[UB]

Pretty Interesting, at least for me. I ran the RC5 Long Benchmark on three of my systems and the results are as follows:

Intel OR840 Motherboard
Dual Slot-1 PIII 667EB
Windows NT 4.0 SP6a
1,917,410.74 KEYS/SEC - 6.9572 KEYS per MHZ per SECOND

ECS D6VAA
Dual Socket-370 PIII 933EB
Windows XP Professional
2,684,275.04 KEYS/SEC - 6.9515 KEYS per MHZ per SECOND

Tyan S2460
Dual Socket-A AMD Athlon 1600+ MP (1400 MHZ)
Windows XP Professional
5,172,332.39 KEYS/SEC - 5.4134 KEYS per MHZ per SECOND

So I could technically calculate 9,774,118.17 KEYS per SECOND. Additionally, the Intel OR840 Motherboard with the dual Slot-1 PIII 667EB processors manages to calculate the most efficient per MHZ, although this could be in relationship to the Operating System. Nevertheless, the Athlon MP setup seems to be the least efficient per MHZ, but easily wins with raw speed.

Any comments or questions?

UB

 

[bot2600]

I think it may have chosen the wrong core for the XP's I have a dual duron 800 setup on a tiger mp motherboard and it gets 5.4 mk/sec so you should be getting MUCH more than that.

Bot

 

[russmanuk]

UB

I think you've made an error in your maths somewhere. The Athlon/Duron series is about 20 faster per Mhz than a P3.

Taking your figures.
P3
2,684,275.04 KEYS/SEC = 2877 KEYS per MHZ per SECOND

MP
5,172,332.39/1400 = 3694 KEYS per MHZ per SECOND

The athlon is 28% faster than the P3

Russman