HPC Warmup Flashcard

Front

The industry standard for programming for NVIDIA GPUs as a accelerator is

Back

CUDA

Front

What is the industry standard for threaded multi-core programming in HPC?

Back

OpenMP

Front

What is the industry standard protocol for distributed computing in HPC?

Back

MPI

Front

In threaded parallelism, when multiple threads can potentially write to the same memory location at the same time, this is called ...

Back

Race Condition

Front

When two or more threads modify different variables that happen to share a cache line, causing unnecessary invalidation and slowdown, this is known as…

Back

False sharing

Front

Amdahl’s Law tells us that speedup is limited by…

Back

The fraction of the code that is sequential

Front

A job with --nodes=1 --ntasks=16 will:

Back

Launch 16 tasks on the same node

Front

For a program that is 95% parallelizable, a reasonable choice for the number of nodes to use is:

Back

20 nodes

Front

The biggest difference between distributed and threaded parallelism is:

Back

Distributed programs communicate by messages; threaded programs share memory

Front

In OpenMP, which variable is always private by default inside parallel for loops?

Back

The loop index