Carnegie Mellon University · 15-418

Parallel Computer Architecture and Programming

并行计算机体系结构与编程

Parallel programming and architecture across multicore processors, GPUs, CUDA, scheduling, locality, cache coherence, interconnects, memory consistency, synchronization, specialization, graph processing, and parallel deep learning.

Start with one small step

Official lecture 1: Why Parallelism

No path choice or account is required. Start one real official item and OpenStudy will remember where you left off.

Start the first lesson ↗Preview a gentle study plan

Subject：Parallel Computing

Level：Advanced Undergraduate

Suggested stage：Years 3–4 (inferred)

Public-material edition：2026

Language：English

Videos：Available

Assignments / solutions：Available / Not verified

Prerequisites

CMU 15-213 or equivalent

Start here

All official course content

Lectures, video, assignments and exams are separated so you do not have to hunt through the course homepage.

36 direct tasks

▶Lectures, video & readingCourse content in the official sequence25

Official lecture 1: Why Parallelism Official lecture 2: Modern Multicore Processors Official lecture 3: Parallel Programming Models Official lecture 4: Parallel Programming Basics Official lecture 5: Work Distribution and Scheduling Official lecture 6: Locality, Communication, and Contention Official lecture 7: GPU Architecture and CUDA Official lecture 8: Parallel Application Case Studies Official lecture 9: Workload-Driven Performance Evaluation Official lecture 10: Interconnection Networks Official lecture 11: Performance Measurement and Tuning Official lecture 12: Snooping Cache Coherence Official lecture 13: Directory-Based Cache Coherence Official lecture 14: Snooping Implementation Official lecture 15: Memory Consistency Official lecture 16: Implementing Synchronization Official lecture 17: Fine-Grained Synchronization and Lock-Free Programming Official lecture 18: Transactional Memory Official lecture 19: Message Passing and Parallel Runtimes Official lecture 20: Prefetching Official lecture 21: Domain-Specific Programming Languages Official lecture 22: Domain-Specific Graph Programming Official lecture 23: Heterogeneous Parallelism Official lecture 24: Parallel Deep Learning: Data Parallelism Official lecture 25: Parallel Deep Learning: Model and Pipeline Parallelism

✓Assignments & practiceOpen the matching problem or assignment page4

Programming assignment 1: Exploring Parallel Computing Programming assignment 2: GPU Programming in CUDA Programming assignment 3: Parallel VLSI Wire Routing with OpenMP Programming assignment 4: Parallel VLSI Wire Routing with MPI

◆Labs & projectsCourse projects, labs and milestones5

Project Proposal Milestone Report Milestone Meeting Final Report Poster Session

◇Exams & quizzesOfficial quizzes, exams and review material2

Exam 1 Exam 2

Official resource hubs

Use these to browse a complete resource collection; the section above links directly to individual study items.

Official lecture schedule, slides, and linked videosCarnegie Mellon University School of Computer Science ↗

Official programming assignment sequenceCarnegie Mellon University School of Computer Science ↗

Official exam dates and review materialsCarnegie Mellon University School of Computer Science ↗

Official project milestones and examplesCarnegie Mellon University School of Computer Science ↗

Source and verification

Every link points to an official university or course-team page. The year identifies the linked public-material edition, not necessarily the latest on-campus offering; a school may have newer materials that are not public. Unknown facts remain unverified rather than guessed.

Carnegie Mellon University School of Computer Science · verified 2026-08-22

Open official course ↗Report a resource issue