TDDD56 Multicore and GPU Programming

Timetable and Lecture Plan

Schedule

Schedule (as available on the LiU schedule server)

Lecture/Lesson Plan

Certain lecture notes and other handouts with restricted access are located here.
The lecture notes and other material may be updated during the course as appropriate.

Lecturers: Christoph Kessler (CK), Ingemar Ragnemalm (IR), Ola Dahl (OD).
Assistants: Nicolas Melot (NM), Lu Li (LL), Jens Ogniewski (JO)

Those lectures marked below in blue color overlap largely with similar lectures in TDDC78 Programming of Parallel Computers - Methods and Tools. These lectures are optional for those who have already taken TDDC78, but might be a useful repetition anyway.

The following list of lecture topics and lessons is preliminary.

• Lecture 1:
  Organization, Overview.
  Motivation: The Multicore Challenge. Multicore Architecture Concepts. (CK)
• Lecture 2:
  Parallel Programming with Threads and Tasks. (CK)
• Lesson 1:
  Introduction to CPU laboratory work. (NM)
• Lecture 3:
  Shared memory architecture concepts and performance issues. (CK)
  
• Lecture 4:
  Theory: Parallel programming and cost models. Analysis of parallel algorithms.
• Lecture 5:
  Theory (cont.): Speedup anomalies. Amdahl's Law. Fundamental parallel algorithms: parallel prefix sums, parallel list ranking. (CK)
  Parallel sorting algorithms: Simple parallel quicksort. (CK)
• Lecture 6:
  Non-blocking synchronization. (CK)
• Lecture 7:
  Parallel sorting algorithms (cont.): Samplesort, Fully parallel quicksort, Bitonic sort, Parallel Mergesort. (CK)
• Lecture 8:
  Parallelization of sequential programs. (CK)
  Mid-term evaluation.
• Lesson 2:
  Selected CPU/theory exercises. (NM)
  Please solve suggested exercises in advance to be prepared.
  
• Lecture 9:
  GPU architecture and trends (IR)
• Lecture 10:
  Introduction to CUDA programming. (IR)
• Lecture 11:
  CUDA programming. GPU lab introduction. (IR)
• Lecture 12:
  Sorting on GPU. Advanced CUDA issues. (IR)
• Lecture 13:
  Introduction to OpenCL. (IR)
• Lecture 14:
  Multicore DSP Architecture and Programming. (OD)
• Lesson 3:
  OpenCL. Shader programming.
  Selected exercises. (IR, JO)
• Lecture 15:
  Parallel Design Patterns and Skeleton Programming.
  Parallel Sorting Contest Awards.
  Autotuning. (CK)

Lab schedule

We have three lab groups, see the schedule.

• Group_A (32 students, room Southfork, supervised by Nicolas Melot and Jens Ogniewski)
• Group_B (16 students, room IDA-Multicore Lab, supervised by Lu Li and Ingemar Ragnemalm)
• Group_C (16 students, room IDA-Multicore Lab, supervised by Nicolas Melot and Jens Ogniewski)

Groups A and B are mostly running in parallel. Group C uses different time slots in the schedule. Register for one of these groups in webreg.

Lab rooms:

• ISY Southfork (Group A)
  
• IDA Multicore Lab ("Konrad Zuse") (B 327:197) (Groups B, C)

Presence in the lab sessions is mandatory.

Deadlines

see the lab page.

Page responsible: Christoph W Kessler
Last updated: 2012-12-06