LearnViz CompSci
Home / Learn / Concurrency & Parallelism / False Sharing

Advanced Topic 41 of 51 Concurrency & Parallelism

False Sharing

Performance loss when threads modify data on one cache line.

How it works

False sharing occurs when threads on different cores modify independent variables that happen to reside on the same cache line, typically 64 bytes. Because cache coherence works at line granularity, each write invalidates the line in other cores' caches, forcing them to reload it, so the line ping-pongs between cores and performance collapses even though the threads share no actual data. It is a performance problem, not a correctness problem.

Where it is used

Per-thread counters in an array, adjacent fields of a struct updated by different threads, statistics arrays and queue head and tail indexes next to each other commonly suffer from it.

Common pitfall

It is hard to notice because the code is logically correct, appearing as poor scaling with more threads. Fix it by padding or aligning hot variables to cache-line boundaries, using alignas or annotations such as Java's Contended, giving each thread its own local accumulators and merging at the end. Profile with hardware counters such as cache misses and tools like perf c2c.

Where it fits in the course

False Sharing is topic 41 of 51 in Concurrency & Parallelism, rated Advanced. It follows Load Balancing in Parallel Programs. The next topic is NUMA.

Practice and explore

How to study this topic

  1. Read the explanation above and restate the definition in your own words.
  2. Open the interactive lesson and answer its mini quiz.
  3. Read the common pitfall, then write a short example that avoids it.
  4. Check your understanding in the Concurrency & Parallelism quiz, and revisit the topic if you miss a question.

Frequently asked questions

What is False Sharing?
Performance loss when threads modify data on one cache line. It belongs to Concurrency & Parallelism and is an advanced-level topic.
Where do people use False Sharing?
Per-thread counters in an array, adjacent fields of a struct updated by different threads, statistics arrays and queue head and tail indexes next to each other commonly suffer from it.
What is the most common mistake with False Sharing?
It is hard to notice because the code is logically correct, appearing as poor scaling with more threads.
What should I learn before this topic?
Start with Load Balancing in Parallel Programs, the topic just before it in Concurrency & Parallelism.
How can I practice it?
Answer the mini quiz in the interactive lesson, then take the Concurrency & Parallelism quiz.
Is False Sharing hard to learn?
It is an Advanced topic. Expect to read the explanation more than once and to connect it with earlier topics. Working through a small example by hand helps most.
How long does it take to learn False Sharing?
Most learners can read the explanation and finish the interactive lesson in under an hour. Fluency comes from applying the idea, so plan a few short practice sessions with the quiz.
Which topics are related to False Sharing?
Load Balancing in Parallel Programs, NUMA, Parallel Sorting are closely connected. The full list is in the related topics section of this page.
What should I study after False Sharing?
Continue with NUMA, the next topic in Concurrency & Parallelism.

Related topics