Live cohortTechnologyLATENCY
Low-Latency C++ for Trading
A 16-week live course that teaches low-latency C++ by building and measuring a matching engine.
Cohort startsDate to be announced
Duration16 weeks · ~10 hrs a week
Curriculum8 modules · 24 lessons
LevelIntermediate
LanguageEnglish
Fee₹50,000
What you will be able to do
Build an order book, matching engine and feed handler with measured p50, p99 and p99.9 latency
Use perf counters and compiler output to explain why code got faster
Use atomics, acquire/release ordering and lock-free ring buffers correctly
Who it is for
- Already writes C++ or knows another systems language
- Comfortable on the Linux command line
- Can commit about 10 hours a week for 16 weeks
- Wants to understand the hardware beneath your code
Curriculum
8 modules, 160 hours
- LAT-01Modern C++ for PerformanceMove semantics and value categories · RAII, templates and constexpr · Hidden allocations in standard library calls20 h
- LAT-02How the CPU Actually Runs Your CodeCache hierarchy and cache lines · Prefetching and pipelines · Branch prediction and the TLB20 h
- LAT-03Memory: Layout, Allocation & PoolsStruct-of-arrays vs array-of-structs · Alignment and object pools · Arena allocators and flat containers20 h
- LAT-04Concurrency & the C++ Memory ModelAtomics and memory ordering · False sharing · SPSC and MPSC ring buffers25 h
- LAT-05Compile-Time TechniquesCRTP and templates · constexpr instead of virtual dispatch · Moving work out of the hot path15 h
- LAT-06Measurement & Profilingrdtsc timing and perf · Google Benchmark · Reading assembly and benchmarking pitfalls15 h
- LAT-07System Tuning & NetworkingCore pinning, isolated CPUs and huge pages · NUMA and busy-polling · UDP multicast and kernel-bypass overview20 h
- LAT-08Capstone: Order Book & Matching EngineLimit order book and matching engine · Feed handler end to end · Profiling, tuning and tail-latency write-up25 h