Resources

Course Resources

• CLRS 4/e - Chapter 26 - Parallel Algorithms
• 10_Tableau_Construction.pdf
• x86 Cheatsheet
• N-Queens Code
• LLVM IR Cheatsheet
• There's Plenty of Room at the Top
• Divide-and-Conquer_Recurrences.pdf
• SchardlDeDo17.pdf
• Master_Method_Business_Cards.pdf
• MIT 6.172: Performance Engineering of Software Systems

External Resources

• CUDA C++ Programming Guide (Link)

  • Table of instruction throughputs (Link)
    (Note: we are using Compute Capability 8.6)

• PTX Reference Manual (Link)

  • Inline PTX syntax guide (Link)

  • Tensor core instruction data layouts (Link)

• SASS Instruction List (Link)

• Compiler Explorer by Matt Godbolt (Link)

  • Lets you view the assembly generated for C++ and CUDA programs
    (among many other languages!)

  • Compiler Explorer page configured for this course (Link)

• GPU Mode (Link)

  • Online reading group and Discord community devoted to studying high-performance GPU programming

Computing Resources

SpeedCode

CSAIL Slurm

Satori

NCSA delta

Engaging