7-Zip Port to CUDA

I want the LGPL-licensed program 7-Zip to be altered to use CUDA for a performance boost. It is already a multi-threaded design so porting it to the GPU this way should be a simple task (relatively), just time consuming. In order to be considered "Complete", this project must run its compression algorithms (all of them) from the GPU only, and if for some reason using a mix between CPU and GPU gives better performance, explain why and provide comparison to prove your point.

## Deliverables

Steps necessary to complete this project and receive payment:

1. Download and extract the source-code provided on this page (also obtainable from [url removed, login to view]

2. Alter any and all files necessary to perform the compression techniques included on the GPU using the CUDA high level programming language provided by nVidia (who's files can be obtained at [url removed, login to view]), optimizing for the most performance possible (even if it means balancing between GPU and CPU or using both to their fullest).

3. Compile and test on a CUDA-capable computer (a PC with one of [these cards][1] installed with the latest CUDA driver) and benchmark against the original 7-Zip on each compression type to determine optimal settings. If this hardware isn't available to you, I'll do it.

Habilidades: Programação C, Engenharia, Linux, Microsoft, MySQL, PHP, Gestão de projetos, Arquitetura de software, Teste de Software, Área de trabalho do Windows

Veja mais: zip cuda, cuda zip, zip gpu, gpu zip compression, cuda compression zip, zip compression cuda, cuda zip compression, porting zip cuda, cuda compression algorithms, gpu cuda compression zip, www c programming language com, www computer programming com, use of algorithms in programming, type of computer programming language, type of algorithms, test algorithms, steps to computer programming, simple algorithms in c, programming with algorithms, programming techniques, programming latest, programming language performance, programming language comparison, nvidia's cuda, nvidia.com

Acerca do Empregador:
( 2 comentários ) Sheppard Air Force Base, United States

ID do Projeto: #3003725