Program Performance & Optimization (Compilers, Loops, Memory, Parallelism)
Автор: Dr. Sam El Gbouri
Загружено: 2026-02-24
Просмотров: 2
Описание:
In this CSC215 Computer Systems lesson (Chapter 6), we break down program performance and optimization in a clear, practical way. You’ll learn how to measure performance correctly, understand what optimizing compilers can (and can’t) do, and apply common strategies that make real programs run faster.
In this video, we cover:
Capabilities & limitations of optimizing compilers (what helps, what blocks optimization)
Expressing performance: execution time, throughput, and speedup
A systematic optimization mindset: measure → change → measure again
Eliminating loop inefficiencies and reducing wasted work
Reducing procedure-call overhead and why it matters in hot paths
Eliminating unneeded memory references (reducing memory traffic)
How modern processors affect performance (pipelines, ILP, stalls)
Loop unrolling and when it helps (and when it hurts)
Enhancing parallelism by reducing dependencies
Memory performance and why caches/locality can dominate runtime
Real-world techniques: bottleneck identification and practical improvements
Повторяем попытку...
Доступные форматы для скачивания:
Скачать видео
-
Информация по загрузке:
