Glossary

Key Points

Introduction	MPI is one of many ways to take advantage of parallel computers. Many different strategies to parallel programming. Use the most appropriate strategy for your problem.
MPI standard	MPI allowed programmers to write portable code across many different supercomputer architectures. MPI is available is available in many languages. MPI requires function calls to pass data around.
MPI point to point communication	Sending messages to another processor is like sending a letter. Non-blocking is most flexible type of point to point communication - just make sure you check for completion.
MPI collective communication	Collective communications make the code easier to understand the purpose.
Advanced topics in MPI	MPI can deliver efficient disk I/O if designed Providing MPI with some knowledge of topology can make MPI do all the hard work. The different ways threads can work with MPI dictates how you can use MPI and OpenMP together.
Summary	MPI is supported in C, Fortran and therefore can be called from many other languages including Python. Just need to know the translation to the other language.

The following list captures terms that need to be added to this glossary. This is a great way to contribute.

Accelerator: to be defined
Beowulf cluster: to be defined
Central processing unit: to be defined
Cloud computing: to be defined
Cluster: a collection of computers configured to enable collaboration on a common task by means of purposefully configured hardware (e.g., networking) and software (e.g. workload management).
Distributed memory: to be defined
Grid computing: to be defined
High availability computing: to be defined
High performance computing: to be defined
Interconnect: to be defined
Node: to be defined
Parallel: to be defined
Serial: to be defined
Server: to be defined
Shared memory: to be defined
Slurm: to be defined
Supercomputer: … “a major scientific instrument” …
Workstation: to be defined
Grid Engine: to be defined
Parallel File System: to be defined