Statements (26)
Predicate | Object |
---|---|
gptkbp:instanceOf |
CUDA API feature
|
gptkbp:canBe |
detecting completion of streams
overlapping computation and communication profiling GPU code |
gptkbp:canBeBlocking |
yes
|
gptkbp:canBeCaptured |
event occurrence on GPU
|
gptkbp:canBeNonBlocking |
yes
|
gptkbp:canSynchronize |
CPU and GPU
|
gptkbp:createdBy |
cudaEventCreate
|
gptkbp:destroyedBy |
cudaEventDestroy
|
gptkbp:documentation |
https://docs.nvidia.com/cuda/cuda-runtime-api/group__CUDART__EVENT.html
|
gptkbp:elapsedTimeMeasuredWith |
cudaEventElapsedTime
|
https://www.w3.org/2000/01/rdf-schema#label |
CUDA Events
|
gptkbp:measures |
elapsed time between events
|
gptkbp:provides |
gptkb:NVIDIA_CUDA
|
gptkbp:queriedWith |
cudaEventQuery
|
gptkbp:recordedAt |
cudaEventRecord
|
gptkbp:scope |
per device
|
gptkbp:syncsWith |
cudaEventSynchronize
|
gptkbp:usedFor |
timing
synchronization |
gptkbp:usedIn |
GPU programming
|
gptkbp:bfsParent |
gptkb:CUDA_Graph_API
gptkb:CUDA_Graphs gptkb:NVIDIA_CUDA_Toolkit |
gptkbp:bfsLayer |
7
|