Colibri T20 Benchmark

Applicable for

colibri t20

Compare with Revision

Subscribe for this article updates

BYTEmark

Note: As far as I can see from the System Monitor the following benchmark only exercised one single core. This Benchmark was done using BSP V1.x.

root@colibri_t20:~# wget http://my.arava.co.il/~matan/770/n900/NNET.DAT
Connecting to my.arava.co.il (94.188.147.3:80)
NNET.DAT             100% |*******************************|  3160  --:--:-- ETA
root@colibri_t20:~# nbench

BYTEmark* Native Mode Benchmark ver. 2 (10/95)
Index-split by Andrew D. Balsa (11/97)
Linux/Unix* port by Uwe F. Mayer (12/96,11/97)

TEST                : Iterations/sec.  : Old Index   : New Index
                    :                  : Pentium 90* : AMD K6/233*
--------------------:------------------:-------------:------------
NUMERIC SORT        :          462.77  :      11.87  :       3.90
STRING SORT         :          57.257  :      25.58  :       3.96
BITFIELD            :      1.3954e+08  :      23.94  :       5.00
FP EMULATION        :          55.116  :      26.45  :       6.10
FOURIER             :          7078.7  :       8.05  :       4.52
ASSIGNMENT          :          7.1372  :      27.16  :       7.04
IDEA                :          1164.2  :      17.81  :       5.29
HUFFMAN             :          809.96  :      22.46  :       7.17
NEURAL NET          :          8.3866  :      13.47  :       5.67
LU DECOMPOSITION    :          293.28  :      15.19  :      10.97
==========================ORIGINAL BYTEMARK RESULTS==========================
INTEGER INDEX       : 21.448
FLOATING-POINT INDEX: 11.811
Baseline (MSDOS*)   : Pentium* 90, 256 KB L2-cache, Watcom* compiler 10.0
==============================LINUX DATA BELOW===============================
CPU                 : Dual
L2 Cache            :
OS                  : Linux 2.6.36.2-svn182
C compiler          : ccache
libc                : static
MEMORY INDEX        : 5.186
INTEGER INDEX       : 5.480
FLOATING-POINT INDEX: 6.551
Baseline (LINUX)    : AMD K6/233*, 512 KB L2-cache, gcc 2.7.2.3, libc-5.4.38
* Trademarks are property of their respective holder.

Ethernet Benchmark

See Ethernet/Network (Linux) - Ethernet Benchmark (iperf).

Floating Point

The Tegra 2 features two Cortex-A9 MPCore ARMv7-A cores with a VFPv3-D16 FPU but does not implement any Neon extension.

Floating Point Benchmark (pi)

With below compiler switches the pi benchmark (http://wiki.toradex.com/index.php/Colibri_Floating_Point_Unit#Performance_Example) can be run on the target in 80 ms:

root@colibri_t20:~# time ./pi
Starting PI...
x= 0.38631 y= 0.89070 low= 939239 j=1200001
Pi =  3.130797 ztot=   801773.75 itot= 1200000
real    0m 0.08s
user    0m 0.08s
sys     0m 0.00s