# Arena Physica’s Heaviside EM model joins Microsoft Discovery for faster design checks

Digest AI · Research · published 2026-10-13T17:53:01Z

Canonical: https://digestai.news/story/arena-physicas-heaviside-em-model-joins-microsoft-discovery-for-faster

## Summary

Arena Physica announced on October 13 that its foundation model for electromagnetism, Heaviside, is now available on Microsoft Discovery, a platform for agentic engineering. The model predicts full 3D electromagnetic field responses in milliseconds—about 100,000 times faster than traditional solvers—addressing what the company calls the ‘verification problem’ in electromagnetic design. Engineers currently spend hours or days running simulations to validate designs, but Heaviside-1, released in early September, cuts that time to near-instantaneous predictions with commercial-grade accuracy (within a decibel of benchmarks). The model, trained on 250,000 3D simulations and 500 billion field samples, supports fully 3D geometries, materials, and excitation inputs, expanding from Heaviside-0’s 2.5D planar focus. Arena Physica introduced EMVal v1.0, a custom benchmark suite, reporting 19% global error on trained data and 33% on unseen geometries, with median vector alignment at 98%. Microsoft Discovery integrates Heaviside with high-performance compute, letting users toggle between AI predictions and traditional solvers for final validation. The platform’s governance controls—audit logging, access controls, and Microsoft 365 integration—are highlighted for handling sensitive IP, such as defense antennas or unreleased chip interconnects. Preeth Chengappa of Microsoft Discovery noted electromagnetic verification remains a bottleneck in hardware design, particularly for signal integrity and thermal co-design. Arena Physica, founded in 2019, partners with companies like AMD and Anduril, and plans to extend Heaviside’s capabilities to far-field radiation patterns in the future.

## Key points

- Heaviside predicts 3D electromagnetic fields in milliseconds, 100,000x faster than traditional solvers, with commercial-grade accuracy
- Model joins Microsoft Discovery’s platform for agentic engineering, integrating with high-performance compute and governance tools
- Arena Physica introduced EMVal v1.0 benchmark, reporting 19% global error on trained data and 33% on unseen geometries

## Why it matters

Faster electromagnetic verification could slash design cycles for semiconductors, defense, and chip packaging, where simulation delays bottleneck innovation.

## Sources

1. [Arena Physica Brings Heaviside EM Model to Microsoft Discovery](https://unite.ai/arena-physica-brings-heaviside-em-model-to-microsoft-discovery) (Unite.AI, 2026-10-13)

## Cite

Digest AI, "Arena Physica’s Heaviside EM model joins Microsoft Discovery for faster design checks", 13 October 2026, https://digestai.news/story/arena-physicas-heaviside-em-model-joins-microsoft-discovery-for-faster

---

Written by Digest AI's editorial model from the linked sources; the sources are the record. Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse
JSON: https://digestai.news/story/arena-physicas-heaviside-em-model-joins-microsoft-discovery-for-faster.json
