Tour of Lux · chapter 8 of 8

Energy as an observable

The part that makes Lux, Lux.

energy.total() reads the meter from inside the program — cost is a first-class observable, queryable mid-run like any other value. In this sandbox it's modeled from evaluation steps (~1 nJ each, and labeled as a model); on real hardware the same call reports measured joules from the attached sensor source.

fn main():
    let before = energy.total()
    let mut acc = 0
    for i in 1..=200:
        acc = acc + i * i
    let after = energy.total()
    println("work result: {acc}")
    println("cost of the loop: {after - before} J")

Once cost is observable, comparing implementations is just code. Both functions below compute the same sum — one walks a loop, one uses the closed form. The receipt difference is the lesson, and it's the whole thesis of Lux in five lines: most compute can be a lookup or a formula, and the meter proves it.

fn sum_loop(n: int) -> int:
    let mut total = 0
    for i in 1..=n:
        total = total + i
    total

fn sum_formula(n: int) -> int:
    n * (n + 1) / 2

fn main():
    let a = energy.total()
    let slow = sum_loop(500)
    let b = energy.total()
    let fast = sum_formula(500)
    let c = energy.total()
    println("loop:    {slow}  cost {b - a} J")
    println("formula: {fast}  cost {c - b} J")

← Apps & views Keep going in the playground →