Tutorial Series · Computational Photonics

Running MEEP on an HPC Cluster

A four-part learning path from computer and cluster fundamentals to MEEP simulations, meta-optics, and reproducible HPC workflows.

FDTD MEEP Slurm MPI Python Linux / HPC

The roadmap below is the planned article architecture. Placeholder pages identify topics reserved for future writing.

Series Roadmap

Part I — Understanding the Machine

Build the basic computer and scientific-computing concepts needed to understand HPC.

I.1

What Is HPC?

Computing, HPC, its purpose, and the problems it is designed to help solve.

I.2

HPC vs. Workstation vs. Server vs. Supercomputer

Compare common computing systems by architecture, purpose, and workload.

I.3

How Does a Computer Process Data?

Programs, instructions, data, memory, storage, and the basic execution path.

I.4

Storage, Memory, Cache, and CPU

A beginner-friendly tour of the memory hierarchy, latency, and bandwidth.

I.5

Inside the CPU

Cores, registers, execution units, control logic, cycles, and I/O.

I.6

Why Scientific Computing Needs HPC

How compute, memory, and runtime demands lead from one machine to a cluster.

Part II — Understanding the Cluster

Learn the general cluster workflow, then relate it to the YZ HPC environment.

II.1

Introducing the YZ HPC Cluster

Introduce YZ HPC and distinguish its site-specific details from general HPC concepts.

II.2

YZ HPC Architecture

Reserve an architecture overview from the user's computer to allocated compute resources.

II.3

The Linux Filesystem on an HPC Cluster

Plan where code, persistent files, temporary data, and simulation outputs belong.

II.4

Connecting to YZ HPC with SSH

Outline the local-to-remote connection and safe authentication practices.

II.5

Login Nodes vs. Compute Nodes

Explain the different roles of login and compute nodes in a scheduled workflow.

II.6

How Do HPC Nodes Communicate?

Introduce node-to-node communication, latency, bandwidth, and data exchange.

II.7

Introduction to Slurm

Reserve an introduction to queues, jobs, partitions, resources, and Slurm commands.

II.8

Writing Your First Slurm Job

Outline a batch script, resource requests, submission, execution, and output.

II.9

Lmod and HPC Software Environments

Reserve space for software modules, versions, dependencies, and environment commands.

II.10

MPI and OpenMP

Compare serial, shared-memory, and distributed-memory parallel execution.

Part III — Understanding MEEP

Connect electromagnetic ideas and numerical methods to MEEP models and results.

III.1

From Maxwell's Equations to FDTD

Trace the path from electromagnetic equations to a discretized time-stepping method.

III.2

The Yee Grid

Reserve an explanation of field placement and spatial and temporal staggering.

III.3

The CFL Stability Condition

Introduce timestep, resolution, and the role of the Courant stability condition.

III.4

Understanding MEEP's Core Objects

Map common MEEP objects to geometry, materials, boundaries, sources, and simulations.

III.5

Your First MEEP Simulation: Vacuum

Reserve the first complete model, run, visualization, and output workflow.

III.6

Fabry-Perot Simulation

Plan a slab example connecting reflection, transmission, interference, and resonance.

III.7

Waveguide Simulation

Reserve a guided-wave model and analysis of modes, index, and field profiles.

III.8

Flux and Electromagnetic Observables

Introduce flux monitors and the interpretation of reflection, transmission, and absorption.

III.9

Photonic Crystals

Reserve periodic structures, Bloch wavevectors, bands, gaps, and defects.

III.10

Mie Scattering

Plan a particle-scattering example and comparison with analytical results.

III.11

Dispersive Materials

Reserve frequency-dependent material models and their role in simulations.

Part IV — Using MEEP for Meta-Optics

Build from MEEP fundamentals toward reproducible meta-optics and metalens workflows.

IV.1

What Is a Meta-Atom?

Introduce a meta-atom as a geometry-dependent element in a metasurface design.

IV.2

Complex Electromagnetic Fields

Reserve amplitude, phase, phasors, and extracting phase from complex fields.

IV.3

Building a Phase Library

Outline a geometry sweep that records transmission and phase for later design.

IV.4

Phase Coverage and Meta-Atom Selection

Reserve phase-range and transmission trade-offs for selecting candidate geometries.

IV.5

Beam Deflector

Plan a device example connecting a spatial phase profile to a deflected beam.

IV.6

Metalens Simulation

Outline the progression from phase library to full-wave metalens evaluation.

IV.7

Running Meta-Optics Parameter Sweeps on HPC

Reserve batch workflows for independent geometry sweeps and output organization.

IV.8

Numerical Convergence

Plan convergence checks for resolution, timestep, boundaries, and domain size.

IV.9

Research Reproduction

Capstone outline for modeling, running, validating, and comparing a published result.

Existing Guides

These articles were published before the four-part roadmap was organized. They remain available as supplementary material:

Updated: