Do you want to learn or get a better understanding of CUDA, OpenCL,OpenMP or MPI? Or all of them?

We offer various on-site training courses about applied numerics and parallel programming in terms of multi-core (OpenMP), NVIDIA GPUs (CUDA, OpenCL), AMD GPUs (OpenCL) as well as multi-node/multi-GPU via MPI (Message Passing Interface). The courses cover:

 

GPU Training (Basics):

  • GPU Architecture Overview
  • Data-parallelism
  • CUDA / OpenCL Models
  • Simple Vector Kernel Code (a walk through)
  • Hands-on session
  • GPU Libraries

 

GPU Training (Advanced):

  • Various algorithms and their parallelism
  • Shared-memory optimization
  • Asynchronous operations
  • Parallel kernel computation (Streams)
  • Multi-GPU programming
  • Debuggers
  • Code Profiling
  • Hands-on session

 

Multi-core CPU Training:

  • Multi-core Architecture
  • Various Algorithms and their parallelism
  • OpenMP Model
  • Hands-on session
  • Debuggers
  • Profiling

 

MPI Training:

  • Multi-node Architecture
  • Various Algorithms and their parallelism
  • MPI Model
  • Synchronous and Asynchronous communication
  • Hands-on
  • Debuggers / Profiling
  • Optimization