Does anyone have an example where they have actually got 8-byte shared memory bandwidth on Kepler? I have a kernel which I suspect is (in places at least) limited by shared memory bandwidth. Having counted all of the shared memory reads/writes and compared with the kernel execution time I’m achieving an average of 540GB/s on a GTX 670 (which has a theoretical maximum of about 1750GB/s). I was wondering if it would be worthwhile trying to use 8-byte transactions. My kernel is mostly convolutions with 3 x 3 x 3 separable filters and it seems like it might be quite similar to the “TTI Reverse Time Migration” case study in “GPU Performance Analysis and Optimization” by Paulius Micikevicius.
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| How to calculate shared memory bandwidth? | 5 | 2517 | June 9, 2012 | |
| Modify the shared memory size into eight-bytes mode | 1 | 445 | March 5, 2019 | |
| Maxwell Shared Memory Bank WIDE | 2 | 981 | February 3, 2016 | |
| Why not full occupancy? | 2 | 982 | November 17, 2012 | |
| Shared memory | 2 | 6860 | April 14, 2011 | |
| Scalability question | 3 | 9128 | June 6, 2009 | |
| What's a reasonable memory bandwidth performance to expect? My current maximum is only around 50 | 1 | 633 | July 27, 2010 | |
| What is the peak bandwidth between the device memory and the GPU on Xavier platform? | 2 | 376 | October 18, 2021 | |
| Host2Device bandwidth, Kepler VS Fermi | 4 | 2094 | July 2, 2012 | |
| memory interface width | 1 | 2884 | June 9, 2010 |