Tous les livres

Professionnel

Applis À propos Coach Connexion Commencer la lecture

Quantitative Finance · Glossaire

Qu'est-ce que « False sharing » ?

Definition 3.5 Low-Latency Software · Chapitre 3 — Memory Hierarchy and Caches

False sharing occurs when two cores repeatedly write different variables that lie in the same cache line: neither shares data with the other, yet every write invalidates the other core’s copy and the line travels between them.

False sharing: two threads, pinned to two CPUs, each incrementing its own atomic counter 20 million times, with the counters on one cache line or on two. Measured on a laptop (Intel Core Ultra 7 155H) under WSL2, no isolated cores. Data: bench_mem.py.
Figure 3.3. False sharing: two threads, pinned to two CPUs, each incrementing its own atomic counter 20 million times, with the counters on one cache line or on two. Measured on a laptop (Intel Core Ultra 7 155H) under WSL2, no isolated cores. Data: bench_mem.py.
Lire dans le chapitre →

Termes liés