Notes on material covered in class. Items marked with a year <2020 have not yet been updated for the Spring 2021 semester. If it's 2022 then this message hasn't been updated either.
Syllabus information and course overview. A quick description of what a GPU does and why it's important.
Basic concepts of performance characterization of programs.
Detailed description of the types of cores used in CPUs and accelerators.
A description of the microarchitecture of the NVIDIA GPUs supporting up to Compute Capability (CC) 9.0
A method of estimating the performance of kernels based on the minimum latency and resources used during the executions fo intervals (sections of code).
Notes on the radix sort program presented in class.
Some notes on the Xeon Phi, includes references.
Some brief notes on Larrabee, Knights Ferry, and Knights Corner. These are examples of manycore GPGPU chips, chips having many simple processor cores coupled with wide vector units.
| David M. Koppelman - koppel@ece.lsu.edu | Modified 7 Mar 2025 18:33 (033 UTC) |
| Provide Website Feedback • Accessibility Statement • Privacy Statement |