Statements (38)
Predicate | Object |
---|---|
gptkbp:instanceOf |
gptkb:software
|
gptkbp:API |
gptkb:Python
gptkb:C++ ONNX Runtime integration |
gptkbp:category |
gptkb:artificial_intelligence
gptkb:software GPU computing |
gptkbp:developer |
gptkb:NVIDIA
|
gptkbp:enables |
low-latency inference
high-throughput inference |
gptkbp:features |
dynamic tensor memory
kernel auto-tuning layer fusion multi-stream execution FP16 support DLA support INT8 support precision calibration |
gptkbp:firstReleased |
2017
|
https://www.w3.org/2000/01/rdf-schema#label |
NVIDIA TensorRT
|
gptkbp:latestReleaseVersion |
8.6.1
|
gptkbp:license |
proprietary
|
gptkbp:officialWebsite |
https://developer.nvidia.com/tensorrt
|
gptkbp:operatingSystem |
gptkb:Windows
gptkb:Linux |
gptkbp:purpose |
deep learning inference optimization
|
gptkbp:supports |
gptkb:CUDA
gptkb:TensorFlow gptkb:NVIDIA_GPUs gptkb:PyTorch gptkb:ONNX |
gptkbp:usedFor |
AI inference acceleration
|
gptkbp:usedIn |
autonomous vehicles
data centers robotics healthcare AI |
gptkbp:bfsParent |
gptkb:NVIDIA_AI
|
gptkbp:bfsLayer |
5
|