2026-03-24 13:40:18 +00:00

4.4 KiB

< Summary | Next >

⏱️ Time Model and Cycles

1. Overview

PROMETEU is a time-oriented system.

Nothing happens "instantly".

Every action consumes measurable time, expressed in execution cycles.

This chapter defines:

  • the system's base clock
  • the concept of a PROMETEU cycle
  • how time is distributed across frames
  • how the CAP (Execution Cap) relates to this model

2. System Base Clock

PROMETEU operates with a fixed clock of 60 Hz.

This means:

  • 60 iterations of the main loop per second
  • each iteration corresponds to one frame
  • the clock is deterministic and stable, regardless of the platform

The clock defines the machine's rhythm, not the amount of mandatory work per frame.


3. The PROMETEU Frame

A PROMETEU frame represents a complete unit of execution.

Each frame conceptually executes the following stages:

FRAME N
──────────────
INPUT
UPDATE
DRAW
AUDIO
SYNC
──────────────

The system guarantees:

  • fixed order
  • predictable repetition
  • absence of "ghost" frames

4. PROMETEU Cycles

4.1 Definition

A PROMETEU cycle is the smallest abstract unit of time in the system.

  • each VM instruction consumes a fixed number of cycles
  • peripheral calls also have a cost
  • the cost is documented and stable

Cycles are:

  • countable
  • comparable
  • independent of real hardware

4.2 Why cycles, and not milliseconds?

PROMETEU does not measure time in milliseconds because:

  • milliseconds vary between platforms
  • real clocks differ
  • jitter and latency hide the real cost

Cycles allow stating:

"This program is more expensive than that one,

regardless of where it runs."


5. Per-Frame Budget

Each frame has a maximum cycle budget.

This budget:

  • is reset every frame
  • does not accumulate unused cycles
  • represents the capacity of the "virtual CPU"

Conceptual example

Frame Budget:10,000cycles

Used:
-Update logic:4,200
-Draw calls:3,100
-Audio:900

Remaining:
1,800cycles


6. Separation between Clock and Work

PROMETEU does not require all logic to run every frame.

The programmer is explicitly encouraged to distribute work over time.

Common examples

  • enemy logic every 2 frames (30 Hz)
  • pathfinding every 4 frames (15 Hz)
  • animations independent of AI
  • timers based on frame count

This reflects real-world practices from:

  • classic consoles
  • embedded systems
  • microcontroller firmware

Not everything needs to happen now.


7. Temporal Distribution as Architecture

PROMETEU treats work distribution over time as an architectural decision.

This means that:

  • code that runs every frame is more expensive
  • distributed code is more efficient
  • optimization is, first and foremost, temporal organization

PROMETEU teaches:

performance doesn't just come from "less code",

but from when the code runs.


8. Execution CAP (Contextual Budget)

8.1 What is the CAP

The CAP defines a set of technical limits associated with a specific context, such as:

  • Game Jams
  • academic evaluations
  • technical challenges

The CAP can define:

  • maximum cycle budget per frame
  • memory limit
  • peripheral call limits

8.2 CAP does not block execution

Fundamental rules:

  • the game always runs
  • the game can always be packaged
  • the game can always be played

The CAP never prevents execution.


8.3 CAP generates evidence, not punishment

When a CAP is active, PROMETEU:

  • measures execution
  • records peaks
  • identifies bottlenecks
  • generates a certification report

This report:

  • does not block the game
  • does not invalidate the delivery
  • documents compliance or non-compliance

9. Time-Based Certification

PROMETEU certification analyzes:

  • average cycle usage per frame
  • maximum peaks
  • problematic frames
  • cost distribution

Example:

Target CAP:PROMETEU-LITE
Frame Budget:5,000cycles

Frame 18231:
Used:5,612cycles❌

Primary cause:
enemy.updateAI():1,012cycles

This certification accompanies the game as a technical artifact.


10. Pedagogical Implications

The time and cycles model allows teaching:

  • execution planning
  • time-oriented architecture
  • technical trade-offs
  • reading real profiles

< Summary | Next >