cppvalleyC++ · Systems · Low Latency · HFT
HFT Core Systems

HFT CORE SYSTEMS

Affinity, topology, SMT, and migration

Lesson 22 of 96Phase 03: Linux determinism
Video lesson coming soon

The written objectives and hands-on lab are available now.

cppvalley on YouTube ↗

Learning objectives

What you should understand

  1. Pin work to CPUs using topology-aware choices
  2. Measure migration and sibling-thread interference
  3. Choose when SMT helps or hurts

Lab

Hands-on exercise

Benchmark pinned, migrating and SMT-contended variants.

Deliverable

Evidence of completion

A CPU-placement diagram with comparative latency distributions.

Prerequisite
EP 21

Phase 03

Linux determinism

Full curriculum
21Scheduler classes and real-time hazards22Affinity, topology, SMT, and migration23CPU isolation as a system configuration24Interrupts, NAPI, softirqs, and NIC queues25Memory locking and prefaulting26Power management, turbo, C-states, and thermal behavior27Finding OS noise with rtla, ftrace, and perf28System calls and context switches are different things29Shared memory IPC with explicit ownership30cgroup v2, cpusets, memory nodes, and least privilege31Asynchronous logging: page cache, direct I/O, and io_uring32Phase gate: reproducible low-jitter host profile