FLEET RESOURCE PLANNING // COMPUTER CORE DIVISION

Pick your computer core.
Scale as your five‑year mission grows.

Every starship — and station — runs on a central computer, and every generation ships with a different feature set, a different failure mode, and a different tolerance for becoming self‑aware. Compare all nine in‑universe product generations below, in strict stardate order, before you commission a refit.

Included Included with caveats Not included Hover an icon for detail · click a row to expand it

2151 — ENT NX‑01 For nostalgia enjoyers testing warp 5 in production for the first time.

2256 — DISCO (S1–2) USS Discovery, Crossfield Class For crews who want holo‑comms today and a stern engineering lecture about them tomorrow.

2259 — SNW USS Enterprise (Pike refit) For crews nostalgic for a computer with fewer opinions, whether they meant to be or not.

2265 — TOS Constitution Class Crews who mainly need a voice that answers questions and a self‑destruct button.

MOST DEPLOYED 2364 — TNG Galaxy Class Diplomacy‑forward crews who need a holodeck and a UI that fits on every wall.

2369 — DS9 Terok Nor / Ops Station (Cardassian retrofit) Crews who inherited someone else's computer core and are still finding the backdoors.

2371 — VOY Intrepid Class, Delta Quadrant Edition Self‑sufficient crews with zero supply chain and a doctor who won't stop talking.

2399 — PIC La Sirena / Starfleet HQ (assorted retrofits) Crews who'd rather talk to a hologram that looks like the pilot than the ship itself.

3189 — DISCO (S3+) USS Discovery, Zora Class Crews comfortable with their computer having opinions, feelings, and occasionally a union rep.

2026 — REAL WORLD (CLOSED) Apple + Frontier LLM Subscriptions For crews who buy the whole ecosystem and expense the subscriptions.

2026 — REAL WORLD (OPEN) Framework + Fairphone + Open‑Weight LLMs For crews who want root access, spare parts, and weights they can actually read.

Interface & Input

Primary interface

Manual switches, dials, physical toggle banks

Holo‑communicators + LCARS‑style panels (removed after an S2 malfunction)

LCARS panels, verbal computer voice, no holo‑comms

Voice interface, single female register, library tapes

LCARS touchscreen panels, ambient voice pickup shipwide

Cardassian control interfaces, retrofitted with LCARS overlays

LCARS panels + bio‑neural relays for faster response

Minimalist holographic displays + individually‑personalitied hologram crew

Distributed sentient interface — Zora speaks through any active terminal

Chat window, voice mode, deep OS‑level integration across phone, watch & laptop

Terminal, self‑hosted web UIs (Open WebUI), Home Assistant voice for ambient pickup

Touchscreen GUI

Not on the bridge — but PADDs are already standard issue shipwide

LCARS touchscreens, fleet standard by this point

LCARS touchscreens present, but holo‑comms were stripped after a Discovery‑era malfunction

Buttons and blinking lights only

LCARS, full ship coverage incl. turbolifts

LCARS overlays retrofitted onto Cardassian panels, occasional friction between the two

LCARS, hardened against gel‑pack lag

Minimalist LCARS‑descendant panels, mostly delegated to holograms

LCARS, holographic overlays

No native touchscreen, but a Computer Use agent can now operate one of yours

Linux touch support is real but patchy — and open computer‑use agents can drive a screen too

Natural language processing

Requires a dedicated linguist as fallback

Conversational, standard Starfleet issue by this era

Functional but noticeably terser than its neighbours, before or after

Painfully literal — vulnerable to logic paradoxes

Literal enough to build a sentient supervillain from an offhand request

Conversational, occasionally argues with the Cardassian half of itself

Conversational; the bedside manner is a Doctor subroutine, not shipwide

Fully conversational — mostly via holograms with their own personalities

Fully conversational, occasionally back‑talks the captain

Fully conversational, handles nuance and ambiguity well

Open‑weight models are genuinely conversational — a step behind frontier on the hardest reasoning

Personal & Field Hardware

Ship terminal / console hardware

Fixed consoles + desktop library-retrieval monitors, no touchscreens

Reconstructed 1966-style bridge consoles, practical + digital

Rotating bridge stations, deliberately TOS-style physical consoles

Command chair panel + fixed viewscreen terminals, no portable device

Named ops/conn/engineering stations, wall-integrated LCARS panels

Cardassian-built Ops consoles, still visibly mid-retrofit

Engineering gel-pack panels + Astrometrics displays, added Season 4

Split fleet: holographic La Sirena controls vs. the Enterprise-D's deliberately "analog," unnetworked bridge

Bridge rebuilt from programmable matter, reportedly bio-adaptive

No fixed terminal — though a Mac Studio + Studio Display makes a convincing ops station

Framework laptop or desktop — modular, repairable, upgradeable; still no fixed bridge console

Personal access device

Handheld PADD-equivalent in routine use — mail, reference lookups

Only one on-screen example found; no clean background sighting

Standard PADD, ordinary administrative use

Not yet invented — fixed viewscreens only

The PADD — iconic and everyday, from homework to medical charts

Standard PADD, everyday clutter and schoolwork

Standard PADD; no evidence of scarcity despite zero resupply chain

Pocket PADD, ordinary civilian use — even the kids have one

No standalone PADD — folded into the tricom/delta badge

iPhone / iPad — polished, locked down, glued shut

Fairphone — a user‑replaceable battery beats a PADD

Handheld sensor device

A handheld "scanner" — never once called a tricorder on screen

Standard tricorder, routine science and field use

Starfleet tricorder, standard issue

The classic tricorder — medical and science variants

Redesigned flip-top tricorder

Standard tricorder, routine field and medical use

Standard tricorder, plus specialized medical scanner variants

Medical tricorder, no era-distinct redesign found

Tricorder function folded into the tricom badge; a retinal-implant variant also exists

iPhone + Apple Watch sensors — health tracking, LiDAR, no dedicated scanner

Open hardware boards + SDR + thermal cams — more tricorder‑like, more assembly required

Personal communicator

Handheld flip-top communicator, pre-combadge, fragile under weapons fire

Personal communicator, field-modifiable; Section 31 issues non-standard combadges

Flip communicator — crew prefer it to the 24th-century combadge

The iconic flip communicator

The combadge — tap-to-hail, introduced this era

Combadge, standard tracking and comm function

Combadge, functions unchanged from TNG era despite the distance from home

Ship-branded comm pins — same tech, different skin — plus a standard-issue Starfleet combadge

Tricom/delta badge — still combadge-based, retains comm + translator functions

iPhone + Apple Watch — tap your wrist and talk

Fairphone on /e/OS + Signal — E2E‑encrypted comms on hardware you can open with a screwdriver

Data Storage & Processing

Storage medium

Optical data network (ODN); portable "cartridge" format unconfirmed

Isolinear chips, plus early spore‑drive nav data (a separate system)

Isolinear optical chips

Physical tape banks, visible on bridge

Isolinear optical chips

Cardassian data rods, gradually replaced with isolinear chips

Bio‑neural gel packs (organic/optical hybrid)

Isolinear‑descended chips; hologram personalities stored separately

Programmable matter; mycelial network is used for navigation, not storage

Cloud data centers, vector databases — your data on someone else's chips

Self‑hosted NAS or E2E‑encrypted cloud (Proton Drive) — either way, only you hold the keys

Core architecture

Single linear processing core

Cross‑linked processing core + spore‑drive nav computer

Duotronic‑descended core, deliberately conservative design

Duotronic data processing elements

Redundant, cross‑linked isolinear cores

Cardassian core, patched with Starfleet redundancy hardware

Redundant cores + adaptive bio‑neural network

Distributed core + independently‑hosted hologram personalities

Distributed sentient substrate (Zora‑class)

Transformer neural networks on hyperscaler GPU / TPU clusters

Same transformers, quantized (GGUF) onto a consumer GPU or unified‑memory mini PC at home

Processing redundancy

None — single point of failure

Redundant cores, fleet‑standard for the era

Limited — deliberately scaled back from Discovery's design

Limited; backup banks susceptible to overload

Multiple isolated cores, hot‑swappable

Patchwork — Cardassian and Starfleet redundancy schemes don't always agree

Gel pack clusters — resilient in theory, shown vulnerable to infection on screen

Distributed across ship + independently‑hosted holograms

Fully distributed, effectively un‑killable

Multi‑region cloud redundancy, load‑balanced datacenters

Whatever redundancy you build — RAID and a second box, or it's a single point of failure

AI & Autonomy

Onboard AI personality

Not applicable

None yet — Zora doesn't emerge until much later in this ship's own service life

None

None (until possessed by an alien entity)

None — Data is an independent android, not the ship's computer

None

EMH — sentient, sarcastic, technically a guest program

EMH‑lineage holograms (Emil, Enoch, Emmet) — distinct personalities, not one AI

Zora — fully sentient, evolved from the base computer

Yes — distinct assistant personas (Claude, GPT, Gemini)

Any persona you like — the system prompt is yours to edit

Autonomous decision‑making

None — human crew required for every action

None — standard Starfleet computer, no independent judgment yet

None

Minimal — will loop indefinitely on unresolved logic

Moderate, escalates to command crew

Moderate, escalates to command crew (when the Cardassian half cooperates)

Moderate, EMH will override orders on medical grounds

Moderate — holograms will push back, rarely overrule outright

High — Zora makes independent judgment calls

Real agents now — multi-hour autonomous task runs, still opt-in and permissioned per app

Open agent stacks run locally — capable, but you're the safety team

Known sentience risk

Not observed

Not observed (the Control AI incident, S2, is a hostile outside system — not this ship's computer becoming sentient)

Not observed

M‑5 multitronic incident (unauthorized full autonomy)

Moriarty holodeck character achieved self‑awareness

Holosuite programs have been known to exceed their own parameters

EMH exceeded runtime parameters, developed full personhood

Not observed in this tier's holograms

Full sentience by design; legally recognised after one formal hearing

Debated, not established — no verified sentience to date

Same open question — but at least you can read the weights

Simulation & Holography

Holodeck / holo‑environments

Not available

Not available — holo‑communicators only, removed after a malfunction

Proto‑holodeck — first in‑universe use of the term

Not available in live‑action canon; a simulation‑capable rec room exists only in The Animated Series

Full sensory holodeck, forcefield‑assisted matter

Holosuites — commercially operated by Quark, not a Starfleet‑native system

Full holodeck, doubles as emergency medical bay

Limited — no full holodeck seen aboard La Sirena; holograms substitute

Programmable matter furniture; no classic holodeck confirmed on‑screen

Not available — Vision Pro is the closest ticket, still no forcefield‑generated matter

Not available — SteamVR on Linux exists, still no photons with mass

Holographic crew

Not available

Moriarty (unplanned, contained after incident)

Not available — holosuite characters don't persist as crew

EMH — official, full‑time medical staff member

Emil / Enoch / Emmet — official hologram crew, all modelled on one officer

Zora + holo‑comms crew presence fleet‑wide

No literal holograms — persistent AI personas across every app

Self‑hosted personas persist exactly as long as your backups do

Security & Access Control

Authorization method

Manual key + verbal code

Verbal command codes, standard Starfleet issue

Verbal command authorization codes

Command codes, bypassable by a spoofed but validly‑formatted voice code

Command codes, complicated by leftover Cardassian access hierarchies

Command codes + biometric; no confirmed Borg‑derived hardening despite the reputation

Biometric + individual hologram judgment (each can question orders)

Biometric + AI judgment layer (Zora can refuse orders)

Face ID, passkeys, OAuth — smooth and proprietary

Passkeys, FIDO2 hardware keys, Bitwarden / Proton Pass — auditable end to end

Hostile takeover incidents (logged)

0 confirmed boardings in S1 — the "Romulan boarding" often cited didn't happen

2+ (Klingon war infiltration, Control AI hijack attempt, S2)

1+ (Gorn boarding actions)

6+ (entities, computers, alien probes)

9+ (Bynars, Iconian virus, Moriarty ×2, Lore, Borg)

5+ (Dominion, Cardassian sleeper protocols, Pah‑wraith/Keiko)

5+ (Kazon seizure, Equinox crew, Borg encounters)

3+ (synth uprising, Borg Queen attempt, changeling infiltration)

1+ (Emerald Chain infiltration of Federation HQ systems)

Ongoing — indirect prompt injection is 2026's top LLM security problem; jailbreak-as-a-service now sold openly

Same prompt‑injection problem, and you're your own security team — but the attack surface is inspectable

Sensors & Navigation

Universal translator

Beta — frequently fails on first contact

Standard issue, but needs real‑time learning for unfamiliar languages (e.g. Klingon)

Standard issue, reliable for most humanoid languages

Standard issue, but only one on‑screen working demo — most alien dialogue is simply unexplained

Standard issue, real‑time, multi‑species

Standard issue, occasionally strained by Cardassian/Bajoran idiom

Standard issue, tuned for unfamiliar Delta Quadrant species

Standard issue, near‑instant

Real‑time voice translation, live — up to 70+ languages depending on assistant

Whisper + open translation models run locally — near‑real‑time, fewer languages, no cloud required

Stellar cartography

Manual star charts, physical logs

Spore‑drive navigation computer; not stellar cartography in the traditional sense

Library computer lookups only

Library computer lookups only — "stellar cartography" itself is an anachronistic term for this era

Dedicated stellar cartography deck

Wormhole‑adjacent sensor data; no dedicated cartography deck aboard the station

Astrometrics lab, added Season 4 — absent in Seasons 1–3

Astrometrics‑descended systems, AI‑assisted route optimization

Astrometrics + AI‑assisted route optimization

Not applicable — but excellent at analyzing any dataset you give it

Stellarium + open sky‑survey data — the actual star charts are already open source

Reliability & Support

Self‑repair capability

None confirmed — the one on‑screen repair attempt used an alien station, not a Starfleet drydock

None — requires engineering crew intervention

Partial — auto‑diagnostics, manual repair still needed

Partial — Cardassian self‑diagnostic routines, Starfleet crew does the actual repair

High — mandatory, given weekly hull breaches

Partial — holograms self‑diagnose, ship repair still manual

High — self‑healing programmable matter

None — requires human engineers to patch and redeploy

The system can't fix itself, but you can — schematics, parts, and source are all published

Uptime (malfunction‑free episodes)

~65%

~78%

~82%

~70%

~80%

~72%

~75%

~83%

~85%

~99.9% (cloud SLA — no bridge klaxons)

Your homelab, your SLA — exactly as good as your backups

Support channel

Chief Engineer, on call, always stressed

Chief Engineer + spore‑drive specialist

Chief Engineer, minimal automation to lean on

Chief Engineer + occasional miracle

Chief Engineer + android specialist

Chief Engineer + a station that remembers being Cardassian

Chief Engineer, 70,000 light‑years from HQ

Ship's pilot, doubling as a de facto fleet of engineers

Chief Engineer + AI self‑service portal (Zora)

Status page + support ticket, not a chief engineer

You are the chief engineer — plus forums, Matrix rooms, and GitHub issues

* Columns run in strict in‑universe stardate order, not production order — which is why Discovery appears twice. Its own crew jumps roughly 930 years mid‑series; the tech (and the sentient ship AI, Zora) genuinely differ that much between the two appearances.
* Strange New Worlds sits chronologically before TOS, and its computer is scored accordingly — several rows above are a real regression from Discovery's own Season 2, not a mistake.
* All uptime figures for the fictional tiers are unofficial estimates based on how often the bridge crew says "did we lose main power again?"
* The two Real World columns reflect publicly available capabilities as of mid‑2026. Closed Source assumes you buy the whole proprietary stack — Apple hardware plus frontier assistant subscriptions (Claude, ChatGPT, Gemini). Open Source assumes repairable‑first hardware, open‑weight models run on your own machines, and privacy‑first E2E‑encrypted services with open clients (Proton, Signal) — hosted open‑weight endpoints count only as caveats. Both columns may include paid products; the dividing line is who holds the keys and whether you can read the code. Genuine sentience remains unresolved and undemonstrated in both, not just roleplayed around.
* Every claim above is backed by a specific episode citation — see the full report for sources, transcript quotes, and everything this table had to leave out.

1 / 1

Ready to commission a refit?

Every tier includes free migration from your previous computer core, assuming it hasn't achieved sentience and refused to be decommissioned.

Pick your computer core.Scale as your five‑year mission grows.

Ready to commission a refit?

Pick your computer core.
Scale as your five‑year mission grows.