[Liblas-devel] Performance Comparison 1.21 C Api vs 1.6b3

[Mateusz Loskot]

Hi,

Calculating average as representative of performance makes little sense.
There are no latencies involved like related to networking, just CPU and RAM
is used.
The best time is the final result.

Also, it's important to try to avoid disk cache influence, especially if the
same file
is read in number of iterations.
On Linux, it's easy to do, just drop all disk caches before reading LAS
file:
http://kernelnewbies.org/Linux_2_6_16
On Windows, ... make N copies of the same LAS file and read different copy
in each iteration.

Best regards,

-----
-- 
Mateusz Loskot
http://mateusz.loskot.net
-- 
View this message in context: http://liblas-developers.431198.n3.nabble.com/Liblas-devel-Performance-Comparison-1-21-C-Api-vs-1-6b3-tp2121578p2130605.html
Sent from the libLAS Developers mailing list archive at Nabble.com.