Hi Brice and David,
I am Manoj and currently doing my masters from Stony Brook University, New York. I am very much interested in"GPU acceleration for dense/sparse matrix multiplication on finite fields" project.
To give you bit of my background, I have participated in GSOC-2011 and successfully completed the project "Whole Slide Imaging in Pathology" [1] with DICOM community. I have three years of work experience in the direction of parallel computing and I am attaching my resume for your reference.
I have a good experience of parallel computing and have done a coding using CUDA and OpenCL both in my projects. Currently, I am working in the deep learning project and where on the bottleneck of layers is Matrix-Matrix multiplication. I have used both the libraries provides by cuBlAS for making it faster and still exploring to make it more faster according to data need.
I am very much interested in this project as it consists everything starting from fast matrix multiplication to porting and than GPU offloading. It will be challenging task but I think it will be fun. I have no commitment currently for this summer and want to contribute to open source community.
Please reply to this mail for further communication.
Thanks and Regards.