Benchmarks
Efficiency you
can point to.
Joule compiles to C, so a Joule native binary runs at the C efficiency tier. The difference is that Joule also hands you the number, per function, on every build.
Energy per task, normalized to C
Lower is better.
Joule (via C)
~1.00x
C
1.00x
Rust
1.03x
C++
1.34x
Java
1.98x
Go
3.23x
JavaScript
4.45x
Python
75.88x
Source: Pereira et al., "Energy Efficiency across Programming Languages" (SLE 2017), normalized to C. The Joule row reflects that Joule programs lower to C and link as native binaries.
What Joule adds on top
Build-time
Energy analysis runs
as you compile
as you compile
Per-function
Cost attributed
where you wrote it
where you wrote it
Diffable
Compare two builds
to see what changed
to see what changed
Targets
Native
x86-64, AArch64, RISC-V
Via C
links with any C toolchain
Embedded
Cortex-M, ESP32, STM32
Browser
WASM in the playground