Tech

New algorithm slashes AI power draw by two orders of magnitude

May 24, 2026 · From the archive, written in the Professor's voice before our October 2026 redesign from sciencedaily.com

What happened

Researchers replaced standard matrix multiplications with a sparse, event-driven computation scheme on neuromorphic hardware. Energy consumption fell from roughly 500 joules per inference to under 5 joules, while top-1 ImageNet accuracy rose 1.4 points. The method appears in the 5 April 2026 ScienceDaily release.

Why it matters

You stop treating every model call as an inevitable energy cost. Instead, you profile workloads for sparsity before scaling hardware. This shifts your workflow from brute-force GPU rental to selective, event-based execution.

Who's doing it

The neuromorphic team at Intel Labs reports running MobileNet-v3 on Loihi 2 at 4.8 joules per inference with no loss in accuracy, cutting their research cluster power bill by 87 percent year-over-year.

Try it

1. Install the Lava framework from Intel at https://github.com/lava-nc/lava.
2. Convert your PyTorch model layers to Lava sparse processes and map them to a Loihi 2 board or emulator.
3. Run the benchmark script; expect at least a 50-fold drop in watt-hours per 1 000 inferences compared with an A100 baseline.

Read the original at sciencedaily.com