parallel.md

Parallel Computing

• 1. Hardware
• 2. Software (Parallel programming)
  • 2.1. Platforms
  • 2.2. Languages
    • 2.2.1. CUDA
      • 2.2.1.1. Learning
      • 2.2.1.2. Utilities

1. Hardware

2. Software (Parallel programming)

• Deep Learning from Scratch to GPU - A series on getting started with parallel programming
• (Article) HPC is more parallel than ever

2.1. Platforms

• Radeon Open Compute - Platform for GPU-Enabled HPC and Ultrascale Computing

2.2. Languages

• Futhark - Futhark is a small programming language designed to be compiled to efficient parallel code. It is a statically typed, data-parallel, and purely functional array language in the ML family, and comes with a heavily optimising ahead-of-time compiler that presently generates either GPU code via CUDA and OpenCL, or multi-threaded CPU code.

2.2.1. CUDA

2.2.1.1. Learning

• CUDA 101
• An Even Easier Introduction to CUDA
• CUDA documentation

2.2.1.2. Utilities

• CUDA Occupancy Calculator - Compute the multiprocessor occupancy of a GPU by a given CUDA kernel