Links
H-Matrices
- HLib
- A sequential H-matrix library with functions for FEM- and BEM-discretisation.
- AHMED
- A (mostly) sequential H-matrix library with functions for BEM-discretisation.
Numerical Libraries
- LAPACK
- A free LAPACK implementation in Fortran with restrictions in multi-threaded environments.
- Intel Math Kernel Library
- A thread-safe LAPACK implementation optimised for Intel processors.
- AMD Core Math Library
- A thread-safe LAPACK implementation optimised for AMD processors.
- CLAPACK (PHI-version)
- A thread-safe LAPACK implementation based on CLAPACK, restricted to the functions used by PHI and with a similar configuration system.
Parallel Communication
- MPICH
- A free MPI implementation with communication via TCP and shared memory.
- LAM
- A free MPI implementation with communication via TCP, shared memory, Myrinet and Infiniband.
- PUB
- A free library for BSP (bulk-synchronous parallel) computations with communication via TCP, MPI and shared memory.
Dynamic Memory Management
- Rmalloc
- A fast, multi-threaded allocator with high parallel efficiency (integrated in PHI).
- PTmalloc/PTmalloc2
- A multi-threaded allocators with very low fragmentation.
- Hoard
- A multi-threaded memory allocator with bounded blowup.