Skip to content
ARIS Join the Waitlist

Choosing a Rig Computer

A rig is the computer that lives at your telescope: it runs your equipment, executes sequences, and keeps shooting after you close the app. Four kinds of machine can be that computer. This page compares them so you can pick before you buy; the rig install guide takes over once you have one.

The four paths at a glance

Section titled “The four paths at a glance”
Rig computerSoftwareWhat you get today
Raspberry Pi 5 (Pi 4 works, slower)Raspberry Pi OS or Ubuntu + the ARIS backendThe full feature set — sequencing, Guardian, weather, SkyNode, overnight stacking, remote access — with GUPPI’s lighter built-in model
NVIDIA Jetson (Orin family)Ubuntu-based JetPack + the ARIS backendEverything the Pi runs, plus GUPPI’s full local model and extra headroom for on-rig processing
Windows mini PC + NINANINA 3.x + the ARIS plug-inA first-class rig for planning, monitoring, capture, and sequence hand-off; a subset of ARIS features today (gap list)
Smart telescope (ZWO Seestar)Nothing to install — the scope is the rigEarly access: joins the fleet next to your other rigs, no rig computer needed

Raspberry Pi — the standard rig

Section titled “Raspberry Pi — the standard rig”

A Raspberry Pi 5 is the recommended rig computer and what most ARIS rigs run. It handles the whole stack — INDI drivers, guiding, plate solving, sequencing, live and overnight deep stacking, Guardian, and remote access — and the installer sets all of it up in one run.

A Pi 4 works too, just slower: driver loading, image transfers, and dual-camera sessions all feel the difference. If you are buying new, buy the Pi 5 — and pair it with a real 5 V / 5 A power supply and an SSD (see what you need).

On the GUPPI front, Pi rigs run the lighter built-in model: the core questions — guiding health, sequence status, the night-journal postmortem — answered from the rig’s live telemetry, entirely on the Pi.

NVIDIA Jetson — the GPU rig

Section titled “NVIDIA Jetson — the GPU rig”

An NVIDIA Jetson (Orin family) runs the same ARIS backend on Ubuntu-based JetPack — same INDI stack, same feature set. What the GPU adds is concrete:

  • GUPPI’s full local model. GPU rigs have the headroom to run a language model alongside capture, so the built-in brain answers with the full model — grounded in live telemetry, private, entirely on your own hardware — where a Pi runs the lighter one.
  • Headroom for on-rig processing. Stacking and solving alongside a busy multi-camera night is where extra compute stops being theoretical.

If GUPPI is part of why you want ARIS, the Jetson is the rig computer to build on.

Windows mini PC + NINA — the plug-in rig

Section titled “Windows mini PC + NINA — the plug-in rig”

If your imaging PC already runs NINA — or you want a Windows machine at the scope — the ARIS NINA plug-in makes it a first-class rig: NINA stays the capture engine, and ARIS adds planning, monitoring, and control from any device. Push a sequence built in ARIS to NINA’s Advanced Sequencer and let NINA run the night.

NINA rigs run a subset of ARIS features today — Guardian, the weather strip, and all-sky data are not available on them yet. See the NINA integration guide for the current gap list and setup.

Smart telescopes — no rig computer at all

Section titled “Smart telescopes — no rig computer at all”

A smart telescope carries its own computer, so this path needs none: a ZWO Seestar joins your fleet directly, in early access. It appears on the same dashboard as your INDI and NINA rigs — the main rig sequencing through the night while the Seestar grabs a second target. See Smart Telescopes for current status.

Any other Linux machine

Section titled “Any other Linux machine”

The catch-all: any 64-bit, Debian-based Linux machine can be an ARIS rig — Debian 12+ or Ubuntu 22.04+, with the installer targeting Raspberry Pi OS and Ubuntu. An old mini PC or NUC running Ubuntu is a perfectly good rig. The Compatibility page has the precise requirements.

What’s next

Section titled “What’s next”