Three instruction NEON float prefix sum. I'd wanted to abuse FCMLA (floating-point complex multiply accumulate) for non-complex arithmetic for so long, and I finally came up with something :)

With two unnecessary multiplies to save one instruction, this may only work out on Apple CPUs, but it's a bit of fun.

(For loops you can broadcast the carried value with vfmaq_laneq_f32(scan, ones, prev, 3) for three multiplies saving two instructions. LLVM fights you on that, though.)

[oops, see reply]

I've had some fun and learned a lot from the "Anthropic's Original Performance Take-Home" optimisation challenge (https://github.com/anthropics/original_performance_takehome)

The old scoreboard (https://www.kerneloptimization.fun/) required Twitter login and got overrun by python/rng-exploit submissions, so here's a new one that requires Mastodon for auth, in case anyone has been playing:

https://vliw-challenge.fly.dev/