Go to file
type-two 5d5bc7dcfb M6: GPU card catalog, carriers, switches, storage, VRAM economy
Compute is componentized: 11 parody GPU cards (Ex-Mining P106, GFX 1060
Dumpster, Radish VII Furnace, RTX 4090 Gigachonk, H100 'The Vault',
B200 BigChungus...) install into carrier slots — Budget PC takes one,
the milk-crate rig four, the half rack eight with rack-only datacenter
cards. Cards carry tok/watts/VRAM/heat individually; total cluster VRAM
must hold the running model version or all token output starves.
Storage devices (SSD shoebox / NVMe hoard / NAS tower) gate training:
no checkpoint space, no progress. Network grows a switch layer with
per-switch throughput caps between devices and uplinks. Shop grouped
by category. Old saves migrate with equivalent card loadouts.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-08-01 11:25:15 +10:00
.claude Tier 1 playable slice: bedroom slop empire 2026-08-01 09:55:55 +10:00
public/assets M6: GPU card catalog, carriers, switches, storage, VRAM economy 2026-08-01 11:25:15 +10:00
src M6: GPU card catalog, carriers, switches, storage, VRAM economy 2026-08-01 11:25:15 +10:00
tools M6: GPU card catalog, carriers, switches, storage, VRAM economy 2026-08-01 11:25:15 +10:00
.gitignore Ignore __pycache__ 2026-08-01 10:05:45 +10:00
CLAUDE.md M6: GPU card catalog, carriers, switches, storage, VRAM economy 2026-08-01 11:25:15 +10:00
index.html Tier 1 playable slice: bedroom slop empire 2026-08-01 09:55:55 +10:00
package-lock.json Tier 1 playable slice: bedroom slop empire 2026-08-01 09:55:55 +10:00
package.json Tier 1 playable slice: bedroom slop empire 2026-08-01 09:55:55 +10:00
README.md Tier 1 playable slice: bedroom slop empire 2026-08-01 09:55:55 +10:00
tsconfig.json Tier 1 playable slice: bedroom slop empire 2026-08-01 09:55:55 +10:00
vite.config.ts Tier 1 playable slice: bedroom slop empire 2026-08-01 09:55:55 +10:00

GigaSlop: Compute Tycoon

Factorio meets Software Inc., wrapped in a satirical sim of the AI slop economy. Start with a budget PC in your bedroom generating $0.40/day of AI spam videos; end up draining the county grid to feed slop-generation clusters.

Stack: TypeScript + Vite + PixiJS v8. The simulation (src/sim/) is a deterministic fixed-tick (10 Hz) module with zero renderer dependencies — the Pixi layer (src/render/) and DOM HUD (src/ui/) just read state.

Run

npm install
npm run dev   # http://localhost:5173

Play (Tier 1 slice)

  • You start with a PC, desk, and router. Press W, click the PC, then the router to wire them — slop production starts.
  • Watch the breaker (1400 W bedroom circuit): trip it and your training checkpoint corrupts.
  • Heat is spatial. Hot PCs throttle at 85°C; fans spread heat around, the window AC deletes it.
  • Uplink caps monetizable tokens — saturated cables mean dropped slop.
  • The TRAIN/INFER slider trades cash now (inference → videos → ad revenue) for model tiers later (higher revenue multipliers).

Art pipeline

All art is generated on the local MODELBEAST farm (tools/mb_gen.py): FLUX for the room shell + hardware sprites (bg-removed, autocropped), and the in-game slop thumbnails are genuine SD/FLUX slop — the game about AI slop is made of actual AI slop. Titles baked to src/data/sloptitles.json.

Layout

src/sim/      balance.ts (all tunables) · catalog.ts (hardware defs)
              state.ts · sim.ts (step, place/wire/sell commands)
src/render/   iso.ts (grid↔screen mapping) · scene.ts (Pixi world)
src/ui/       hud.ts + hud.css (DOM dashboards, shop, SlopTube feed)
tools/        mb_gen.py (MODELBEAST batch art generation)