Expert · Module 14 · Lesson 1 of 4

Caches and False Sharing: Layout for Speed

Keep hot data on its own cache line, split counters per thread, and prove the speedup with the honest benchmark harness.

ProAbout 11 minutes

By the end you can

Explain cache lines and why two counters can slow each other down
Use alignas and padding to separate hot fields
Fix a false-sharing slowdown with per-thread sharding

Builds on Benchmarking: Clocks, Warmup and Honest Numbers.

01 · A question first

Eight threads bump eight counters two hundred million times. The counters are independent, no lock, no atomic contention on paper. Yet the loop takes six times longer than one thread doing all the work. What are the threads fighting over if they never touch each other's data?

Make a guess before you read on.

02 · The idea

Members only

Unlock the full lesson

This lesson is part of the membership. Join to read it in full, along with every other premium deep dive.

What's inside

Start with the problem
How the idea works
Why it is correct
What it costs
A sharded counter with separated elements
Pitfalls and variations
Where it is used