Profile and optimize an existing C++ library for speed.
This library uses OpenCV and OpenMP and C++ standard libraries.
This job will require understanding the algorithm and experience in OpenCV, C++ to optimize the algorithm
(i.e. for instance, deciding whether to use matrix algebra vs using an C++ array).
An understanding of parallelization approaches using OpenMP and/or other linear algebra libraries is also required.
This library must work in a unix based system.
C++ programmer with experience in parallel programming using OpenMP.
I have done many optimizations on C/C++ programs.
I can run the library with perf or other Linux tools and find the bottlenecks
where we can do some modifications.
hi
WE are expert coders of c++ and this process can be done about optimization.
good team with experience .
will like to work on this if given chance will share all remaining details regarding your project once awarded.
hope to hear from you
Thank you
Regards
Hi,
Reducing the number of execution, Data caching and Scaling for parallel processing are number of ways of optimization for speed. You know these days parallel processing are getting popular. I you go for GPGPU/CUDA technology of processing you can get speed of more than 20 times for large throughput. As you mentioned this technique will be very fruitful for matrix algebra because every matrix elements can be scaled to compute for individual.
Please contact if you are looking for this kind of speed up.
Regards