Glossary of Key Terms

Term Definition
HPC High Performance Computing (see Primer)
Cluster Servers, networked to execute batch jobs that require high core count, not possible on a single physical server
Scheduler Software responsible for the job, queue, and user management and accounting, e.g. Slurm
Slurm The scheduler used at Pomona HPC
GPU Graphics Processing Unit
GPGPU General Processing Graphics Processing Unit
CUDA Proprietary NVIDIA programming language for highly parallel tasks on GPUs
AI Artificial Intelligence – an all-encompassing term largely meaning computer’s ability to augment human tasks
ML Machine Learning – a subset of AI where a computer can be trained on a specific dataset, supervised or unsupervised, to perform a specific task, sometimes using neural networks
Inference Computer’s ability to infer specific qualities in an object like an image after having been trained on a specific (usually curated by a human) dataset (e.g. recognizing cats in pictures or tagging a specific face in a photograph)
Python A very widely used programming language, especially in Data Science or by people learning their first programming language, interpreted vs compiled (e.g. C)
DevOps A practice of describing systems administration tasks in code to make them scalable and repeatable
Infrastructure As Code Same as DevOps
Configuration Management Same as DevOps
Ansible A DevOps tool, similar to Puppet, Salt, Chef