Notes

Completed

How I Think About Strong Scaling in MPI Workloads

A short note on how I read strong scaling results, and why efficiency and communication cost matter as much as runtime.

Read Note

A Beginner's Handbook to LLM Applications

Tokens, networking, GPU memory, RAG, agents and inference explained with analogies, figures and measured data, ending with a system-design question.

Read Note 中文版

Notes

In Progress

PETSc Benchmarking on ARCHER2

A future note on what performance measurements revealed across MPI-only and hybrid configurations.

Notes

Planned

Weekly Learning Log

A lightweight recurring note format for recording experiments, problems, and next steps.