[Don't Merge]add stream type to CUDAPlace #7986
Closed
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters. Learn more about bidirectional Unicode characters
QiJune reviewed Jan 31, 2018
| }; | ||
| | ||
| struct CUDAPlace { | ||
| enum StreamType { kCOMPUTATION, kCOMMUNICATION, kLast }; |
There was a problem hiding this comment.
Our original idea of Place is to represent the memory location of a tensor. I am not sure it's a good idea to add cuda stream into Place.
There was a problem hiding this comment.
We need one to one mapping between CUDAPlace and CUDAContext. If we have CUDAPlace mapped to multiple CUDACtx, how should the API of DeviceContextPool::Get(const platform::Place& place) look like?
tonyyang-svail changed the title added 3 commits February 6, 2018 00:32
| The goal of the design #7814 is done at ParallelExecutor |
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit. This suggestion is invalid because no changes were made to the code. Suggestions cannot be applied while the pull request is closed. Suggestions cannot be applied while viewing a subset of changes. Only one suggestion per line can be applied in a batch. Add this suggestion to a batch that can be applied as a single commit. Applying suggestions on deleted lines is not supported. You must change the existing code in this line in order to create a valid suggestion. Outdated suggestions cannot be applied. This suggestion has been applied or marked resolved. Suggestions cannot be applied from pending reviews. Suggestions cannot be applied on multi-line comments. Suggestions cannot be applied while the pull request is queued to merge. Suggestion cannot be applied right now. Please check back later.
As part of the design #7814, this PR would allow us to create multiple device contexts on one GPU.