Early Adopter: Integration of Parallel Topics into the Undergraduate CS Curriculum at Calvin College

1
Early AdopterIntegration of Parallel Topics
into the Undergraduate CS Curriculum at Calvin
College

• Joel C. Adams
• Chair, Department of Computer Science
• Calvin College

2
Calvin College

• Is a 4-year comprehensive liberal arts college
  located in Grand Rapids, MI
• Has about 4,000 students
• Offers ABET-accredited programs in
• Computer Science (BCS)
• 7.5 FTE faculty
• 100 students
• Engineering (BSE)

3
CS Major Curriculum
Year
Fall Semester
Spring Semester
1
Intro to Computing Calculus I
Data Structures Calculus II
2
Algorithms DS Intro. Comp. Arch. Discrete Math I
Programming Lang. Discrete Math II
3
Software Engr Adv. Elective
OS Networking Statistics
4
Adv. Elective Sr. Practicum I
Adv. Elective Sr. Practicum II Perspectives on
Comp.
4
CS 112 Intro to Data Structures

• Year 1, semester 2, core, 40 students
• Topics
• Single threading vs. multithreading
• Parallelizing linear search, mergesort
• Race conditions non-thread-safe structures
• Speedup, asymptotic performance analysis
• Hands on Embarrassingly parallel matrix
  operations (addition, transpose) via OpenMP
• Reduced Coverage of graphs

5
CS 212 D. Structures Algorithms

• Year 2, semester 1, core, 12-15 students
• Topics
• Parallel vs distributed algorithms
• Parallel design (data vs task decomposition)
• Parallel algorithms (parallel sorting, max)
• Distributed graph algorithms
• Asymptotic performance analysis
• Reduced Coverage / review of elementary data
  structures (stacks, queues, sets)

6
Engr 220 Intro. Architecture

• Year 2, semester 1, core, 12-15 students
• Topics
• Flynn taxonomy (SISD, SIMD, MIMD)
• Multiprocessors (shared vs distributed mem.)
• Multicore, manycore processor architectures
• Memory hierarchy
• Multicore memory bandwidth issues
• Multicore caching, coherency issues
• Reduced Assembly language coverage

7
CS 214 Programming Languages

• Year 2, semester 2, core, 20 students
• Topics
• Concurrency and synchronization using
• Processes / threads / tasks
• Semaphores, locks, condition variables, monitors
• Race conditions
• Communication constructs
• Message passing vs. shared memory
• Hands on Threads in C, Ruby Ada tasks
• Reduced coverage of markup languages

8
CS 232 OS Networking

• Year 3, semester 2, core, 15 students
• Topics
• Concurrency synchronization implementing
• Processes, threads,
• Semaphores, locks, condition variables, monitors
• Race conditions, deadlocks
• Traditional vs. multicore scheduling
• TCP/IP and Ethernet Networking
• Sockets, message passing, RPC,
• Classic distributed systems problems

9
CS 232 OS Networking (ii)

• Hands on
• Multiprocessing (fork), synchronization
• Multithreading (POSIX), synchronization
• Kernel programming
• Socket programming (client-server systems)
• Reduced coverage of
• Hardware review (from 1 week to 1 lecture)
• Batch systems
• I/O devices

10
CS 374 High Perf. Computing

• Year 3 4, semester 1, elective, 8 students
• Topics
• Multiprocessor architectures
• Problem decomposition data vs task
• Parallel patterns models
• Problem types Monte Carlo, N-body, etc.
• Scalability
• Speedup, Efficiency, Amdahls L., Gustafsons L.
• Hands on MPI, OpenMP, MapReduce, OpenCL

11
Summary

• By sprinkling parallel topics throughout the
  core courses, students
• Are exposed to parallel thinking early (year 1)
• Are exposed to parallel thinking frequently
• Are gradually exposed to increasingly difficult
  aspects of parallelism
• This provides a framework for gradually
  increasing the parallel content as faculty become
  comfortable teaching it.