Hi,
LDGSTS is a instruction about Asynchronous Global to Shared Memcopy. However, I cannot find in any document about cycles about this instruction in A100. I was wondering if there is any reference document about cycles of the ldgsts instruction?
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| Will ldgsts execute concurrently with instructions that load from shared memory? | 1 | 214 | March 12, 2024 | |
| How memcpy_async be asynchronous? | 2 | 306 | June 13, 2024 | |
| CUDA PTX cp.async only supports global to shared memory copy | 2 | 965 | March 14, 2023 | |
| Why Reg->shared->global is faster than Reg->global? | 7 | 1182 | June 11, 2022 | |
| What is the benefit from LDG or LDG128? | 3 | 621 | January 17, 2024 | |
| global memory access synchronous or asynchronous read/write? | 3 | 3368 | May 15, 2008 | |
| Question about Global memory and Texture memory | 5 | 956 | October 23, 2014 | |
| In V100 GPU or A100 GPU, CUDA COREs- data miss in registers where do the CUDA core look first for data - in Shared Memory or in L1 cache? | 4 | 521 | September 27, 2023 | |
| Device to device copy = SLI copy? SLI copy feature? when? | 2 | 3582 | October 14, 2007 | |
| What is the functionality of LD/ST units in SM? | 4 | 407 | May 23, 2024 |