port of MAGMA / MATRICES The University / Sandy Bridge / /
IndustryTerm
possible time to accurate solution / linear algebra reliant software components / parallel computing / vector product / given energy constraints / communication-avoiding algorithms / heterogeneous systems / e-values / dynamic runtime systems / e-vectors / /
OperatingSystem
Mac OS / Linux / Microsoft Windows / /
Organization
LA CPU / 32K 34K 36K Windows MATRIX SIZE Mac OS INDUSTRY COLLABORATION NVIDIA’s CUDA Center of Excellence Program / University of Florida / Department of Defense National Science Foundation FIND OUT MORE AT / U.S. Department of Defense / U.S. Department of Defense National Science Foundation / /
Person
Compute / BLAS CPU / /
Position
DRIVER / STANDARD EVP GE GENERALIZED EVP / MP / using MP / Inverse Iteration MP / SVP / General / /
Product
Sony Ericsson P1i Smartphone / /
SportsEvent
ufl / /
Technology
Linux / hybridization / PKI / communication-avoiding algorithms / /
MAGMA (Matrix Algebra on GPU and Multicore Architectures) is a collection of next generation linear algebra libraries for heterogeneous architectures. MAGMA is designed and implemented by the team that developed LAPACK a