Electromagnetic precision at unprecedented scale.

A new generation of high-order electromagnetic simulation, designed to solve larger problems at higher accuracy and dramatically lower computational cost.

RECOGNITION & AFFILIATIONS

Forge 2026Finalist

Larger problems, linear scaling.

FDTD and low-order FEM force a trade-off between model size, accuracy, and speed. Planck changes that equation.

  1. 01
    O(N) scaling

    Larger problems become attainable.

  2. 02
    100×–1,000× more accurate. Faster.

    High order cuts the degrees of freedom needed.

  3. 03
    Parallel by design

    Distribute the work across GPUs.

Runtime and accuracy

Illustrative
2.00×
PlanckLow-order FEMFDTD

Runtime

Relative runtime (×)0501001502002500×9×18×27×O(N)DOF growth (×)3.6×speedupvs FDTD

Accuracy

Relative error · log scale110⁻¹10⁻²10⁻³10⁻⁴0×9×18×27×DOF growth (×)400×more accuratevs FDTD

Where other technologies cross ours.

Explore the devices and ideas shaping photonics, and the electrical, mechanical, and optical physics behind them.

Browse deep dives

Built by researchers, for device teams.

Planck Labs is a small team building a new kind of electromagnetic solver. Hirish Chandrasekaran is a PhD student in Computational Science at University of Illinois Urbana-Champaign with a strong background in computational science and physics. Kennard Peters is a software engineer with experience in full-stack development.

Explore the solver approach on the , and the deep dives are where we feature device examples and case studies.

FOUNDERS
Hirish Chandrasekaran, PhD Student, University of Illinois Urbana-Champaign
Kennard Peters, Software Engineer
BASED IN
Champaign, IL

Bring a representative device problem.

Geometry, material stack, frequency or wavelength band, and target response are enough to start a conversation.

Discuss a simulation problem

© 2026 Planck Labs · High-order electromagnetic solvers