# Prefix sums at tens of gigabytes per second with ARM NEON

DevFeed: [Prefix sums at tens of gigabytes per second with ARM NEON](<https://devfeed.tech/articles/prefix-sums-at-tens-of-gigabytes-per-second-with-arm-neon-29395.md>)

Original publisher: [Read original article](<https://lemire.me/blog/2026/03/08/prefix-sums-at-tens-of-gigabytes-per-second-with-arm-neon/>)

Author: Daniel Lemire

Published: 2026-03-08T20:09:48Z

Content type: tutorial

Language: en

Sources: [Daniel Lemire](<https://devfeed.tech/sources/daniel-lemire.md>)

Topics: [systems](<https://devfeed.tech/topics/systems.md>), [Arm](<https://devfeed.tech/topics/arm.md>), [cpu](<https://devfeed.tech/topics/cpu.md>), [C](<https://devfeed.tech/topics/c.md>)

Tags: [arm](<https://devfeed.tech/tags/arm.md>), [c](<https://devfeed.tech/tags/c.md>), [cpu](<https://devfeed.tech/tags/cpu.md>), [processor](<https://devfeed.tech/tags/processor.md>)

## AI overview

The article explains how prefix sums can be optimized with ARM NEON SIMD instructions. It compares scalar and vector approaches and describes processing interleaved blocks of 16 integers in parallel to reduce the sequential instruction cost.

## Source excerpt

Suppose that you have a record of your sales per day. You might want to get a running record where, for each day, you are told how many sales you have made since the start of the year. day sales per day running sales 1 10$ 10 $ 2 15$ 25 $ 3 5$ 30 ... Continue reading Prefix sums at tens of gigabytes per second with ARM NEON