|
Boost : |
Subject: Re: [boost] [Countertree + Suballocator] New Version
From: Francisco José Tapia (fjtapia_at_[hidden])
Date: 2012-04-16 17:00:47
Hi, Mr Gaunard , I had forgot all. Smile, and start again.
I didn't wrote a description of the algorithm because I didn't know if
somebody was interested.
Please let me 2 or 3 days, and I will prepare a pdf document with the
description of the algorithm. With the explanation and the diagrams, it is
very easy understand and see the characteristics of the suballocator
described
IMPORTANT . About the benchmarks. There are important differences between
the times obtained in different processors, depending, mainly of the cache
of the processor. The others characteristics as memory allocation don't
change. The suballocator shows their utility when you have many elements
(several millions). With a small number of elements, the results are
similar to the standard allocator.
If somebody have any idea or detect a fail, please, contact me, and help me
to improve the library and make it useful for the boost community in the
same way than others had been useful for us
Sincerely yours
Francisco Tapia
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk