eic / eic/eic-shell

ARM architecture plan

Open
#33 2 comments 0 reactions 0 assignees View on GitHub
Dominant language
Shell
Stars
2
Forks
5
PR merge metrics
No merged PRs in 30d

Description

**Is your feature request related to a problem? Please describe.**

eic-shell currently runs only through AMD64/x86_64 container images. On Apple Silicon systems (ARM64/aarch64), this requires emulation (Rosetta/QEMU), which significantly reduces performance during CPU-intensive workflows such as npsim and especially eicrecon.

For large-scale reconstruction campaigns with background mixing enabled, the slowdown becomes substantial and limits usability on modern ARM64 laptops and workstations.

I understand that much of the HEP and detector simulation ecosystem has historically targeted AMD64/x86_64 platforms, and that some dependencies may still be architecture-specific. However, it may already be possible to support at least part of the software stack natively on ARM64.

**Describe the solution you'd like**

Provide native ARM64/aarch64 support for eic-shell, including:
- ARM64 container images
- multi-architecture Docker builds (linux/amd64 and linux/arm64)
- CI validation on Apple Silicon and/or ARM64 Linux
- documentation for ARM-native installation and execution

Ideally, users on Apple Silicon Macs should be able to run eic-shell natively without requiring x86 emulation.

Even partial ARM64 compatibility for selected components would already be very valuable.

**Describe alternatives you've considered**

Current alternatives include:
- running through Rosetta/QEMU emulation on macOS
- using remote x86_64 servers or clusters
- manually rebuilding dependencies for ARM64

However, these approaches either introduce large performance penalties or significantly increase setup complexity.

**Additional context**

On Apple Silicon systems, workflows involving:
- background overlay,
- npsim,
- and especially eicrecon

can become substantially slower under x86 emulation.

Native ARM64 support would improve:
- developer experience,
- portability,
- battery efficiency,
- and local reconstruction performance for EIC software users on modern hardware platforms.

Contributor guide

No contributing guide indexed for this repository

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.