From 67d7367734fecbed5db91750e3ef02462384326a Mon Sep 17 00:00:00 2001 From: Shane Snyder Date: Tue, 13 Jan 2026 13:06:19 -0600 Subject: [PATCH] update README to account for dev API -D,-V options --- README.md | 2 ++ 1 file changed, 2 insertions(+) diff --git a/README.md b/README.md index b9185a7..ef38c34 100644 --- a/README.md +++ b/README.md @@ -80,6 +80,8 @@ All tests support the same set of arguments : * `-G,--cudagraph ` Capture iterations as a CUDA graph and then replay specified number of times. Default : 0. * `-C,--report_cputime <0/1>` Report CPU time instead of latency. Default : 0. * `-R,--local_register <0/1/2>` enable local (1) or symmetric (2) buffer registration on send/recv buffers. Default : 0. + * `-D,--device_implementation ` use custom device API implementation. Not every collective has a custom device API implementations (currently just all\_reduce and alltoall). Default : 0 (use traditional NCCL host implementation). Note: values > 0 require symmetric memory registration (`-R 2`). + * `-V,--device_cta_count ` number of CTAs for device API implementation. Must be positive and less than 128. Default : 16. * `-S,--report_timestamps <0/1>` Add timestamp (`"%Y-%m-%d %H:%M:%S"`) to each performance report line. Default : 0. * `-J,--output_file ` Write [JSON] output to filepath. Infer type from suffix (only `json` supported presently). * `-T,--timeout