Project announcement · Active R&D · Rust systems
The Generalized Execution Platform is a domain-blind operating layer for scientific research and deployment. Author your mathematics, laws, models, and experiments once. At run time, the engine stitches them together, allocates real machines, moves your data, and executes your work — while your meaning stays exactly yours.
01 / The idea
A game designer never writes driver code. They author objects, textures, shaders, and rules — and when the moment comes, the engine stitches everything together in real time. Science has never had that machine. GEP is being built to be exactly that.
meshes · textures · shaders · physics settings
mathematics · laws · models · methods · schemas · experiments
scenes assembled at runtime by the engine
exact dependency closures, lowered deterministically at activation
rendering · physics · memory · platform targets
execution · data movement · allocation · custody · lineage
drivers · draw calls · allocators
schedulers · kernels · backends · queues · clusters
The platform ships blank — no preferred mathematics, no built-in science, no opinion about your domain. You bring every ounce of meaning.
02 / The steel boundary
The defining cut of GEP is a boundary we treat like steel. Authored meaning lives above it; execution, data plumbing, and physical hardware live below it. Neither side leaks into the other — so either side can evolve without rewriting the other.
Above the line — yours
Below the line — the engine’s
Swap a kernel. Add a GPU class. Move to a cluster. Your science does not change by one byte.
Science on GEP stacks like floors in a building. Each layer is an inert, versioned asset that composes on the exact closure beneath it. Build a tower yourself — click the assets below.
Select an asset to begin stacking.
03 / Architecture
GEP is four peer pillars joined by typed crossings. Each owns a narrow slice absolutely — and is forbidden from reaching into its neighbors’ concerns.
P·1
Authored meaning made exact: versioned definitions, precise dependencies, deterministic lowering, and a separate semantic trace.
P·2
A meaning-blind executor. It runs lowered programs as retained Jobs and returns exact technical outputs — understanding nothing of what they mean.
P·3
Finite, governed data: bounded exchange with custody today; persistence, cataloging, query, and observation as it matures.
P·4
The sole meaning-blind physical governor. It joins lawful work to current machinery through fit, grants, waits, and release.
04 / The route
01
Define meaning and exact dependencies — once.
02
Resolve one exact closure; lower it to a mechanical program plus a separate semantic trace.
03
Bounded technical inputs cross the Data Plane under custody.
04
One retained, meaning-blind Player Job runs.
05
Allocation grants, accounts for, and releases the current physical work.
06
Mechanical evidence and semantic provenance stay distinct — but exactly related, forever.
Because every crossing is typed and every identity exact, the same route serves a two-second script or a month-long campaign — and replays honestly years later.
05 / What gets built on top
No science lives inside the engine — which is why every science can live on top of it. Domains become asset packs: authored once, locked, reused everywhere.
First proving ground: an end-to-end quantitative-trading research vertical. Generality is earned through real executable work — not speculative abstraction.
06 / What’s real today
This is an announcement of direction, not a finished product — so this page doesn’t hand-type its claims. The ledger below reads the project’s own PROJECT_STATUS.md, straight from the working tree. When engineering advances, this page changes with it.
Foundation ledger
Explicitly not yet
Sequenced next
Governance can prove what ran, under what authority, from which exact inputs. It cannot make weak science true.
07 / Operating principles
01
No math, laws, models, or methods ship inside. The platform holds no scientific opinion.
02
Inert, versioned assets composed with exact dependencies belong entirely to their author.
03
Kernels, schedulers, backends, and capacity are platform concerns — never labels in science.
04
One Job’s wait, failure, or long life never stops unrelated compatible work.
05
A composition may branch into an independently owned executable — research never stops.
06
The engine proves provenance and authority. It cannot declare a conclusion scientifically true.