Subject: Re: Benchmarks
To: Ale Terlevich <A.I.Terlevich@durham.ac.uk>
From: Peter Berg <Peter.Berg@ww.tu-freiberg.de>
List: port-arm32
Date: 11/11/1996 13:09:31
<fixed>> Here's a side by side comparison of the two systems.
>
> Arm7/40 SA110/200
>
> Pipe-based Context Switching Test 644.8 lps 372.2 lps
> Process Creation Test 47.0 lps 16.4 lps
> Execl Throughput Test 50.7 lps 16.6 lps
> C Compiler Test 16.3 lpm 13.9 lpm
I must say I'm a little bit disappointed ! The most I do on my system is to
compile some stuff, and now it should be slower then at the moment it is on my
systen ?!
> Shell scripts (1 concurrent) 31.6 lpm 23.3 lpm
> Shell scripts (2 concurrent) 13.0 lpm 10.3 lpm
> Shell scripts (4 concurrent) 6.0 lpm 4.0 lpm
> Shell scripts (8 concurrent) 2.6 lpm 1.0 lpm
> Dc: sqrt(2) to 99 decimal places 1234.6 lpm 462.1 lpm
Can someone explain me why all these tests are slower on a SA110 then on my
ARM710 ?
The next question:
Why were floatingpoint operations in *all* kernels newer then #3854 only half
as fast as in kernel #3854 ?
Here are the results for float with this kernel ?
#3854:
Arithmetic Test (type = float) 212.3 lps (10 secs, 6 samples)
Arithmetic Test (type = double) 202.8 lps (10 secs, 6 samples)
#4193:
Arithmetic Test (type = float) 106.9 lps (10 secs, 6 samples)
Arithmetic Test (type = double) 109.9 lps (10 secs, 6 samples)
</fixed>
Ciao, Peter
_____ _ ____ _____ _____
| __ \(_) | _ \ / ____| __ \
| |__) |_ ___ ___| |_) | (___ | | | |
| _ /| / __|/ __| _ < \___ \| | | |
| | \ \| \__ \ (__| |_) |____) | |__| |
|_| \_\_|___/\___|____/|_____/|_____/
E-Mail: Peter.Berg@ww.tu-freiberg.de
WWW : http://www.ww.tu-freiberg.de/~pberg