cuDNN
E209958
cuDNN is NVIDIA’s GPU-accelerated library of optimized primitives for deep neural networks, widely used to speed up training and inference in frameworks like TensorFlow and PyTorch.
All labels observed (3)
| Label | Occurrences |
|---|---|
| cuDNN canonical | 15 |
| NVIDIA cuDNN | 7 |
| CUDA Deep Neural Network library | 2 |
How this entity was disambiguated
This entity first appeared as the object of triple T1893378 — resolving that mention is where its identity was fixed. The disambiguator weighed these candidate entities and picked the highlighted one (or “None”, minting a new entity). This is how homonymy is resolved: the same surface form can point to different entities.
NED1
Entity disambiguation (via context triple)
gpt-5-mini-2025-08-07
Target entity: cuDNN Context triple: [NVIDIA CUDA, includes, cuDNN]
-
A.
NVIDIA CUDA
NVIDIA CUDA is a parallel computing platform and programming model that enables developers to use NVIDIA GPUs for general-purpose high-performance computing.
-
B.
CuPy
CuPy is an open-source array library for Python that accelerates numerical computing by providing a NumPy-compatible interface backed by GPU execution.
-
C.
PlaidML
PlaidML is an open-source, hardware-agnostic deep learning engine designed to accelerate neural network computation on a wide range of GPUs and other devices.
-
D.
NVIDIA AI Enterprise software suite
NVIDIA AI Enterprise software suite is a comprehensive, enterprise-grade collection of AI tools, frameworks, and optimized software designed to accelerate the development and deployment of AI and data analytics workloads across modern data centers and clouds.
-
E.
Microsoft Cognitive Toolkit
Microsoft Cognitive Toolkit (CNTK) is an open-source deep learning framework developed by Microsoft for building, training, and deploying neural networks at scale.
- F. None of above. chosen
- G. Unsure - the case is ambiguous/there is not enough information to decide.
NED2
Entity disambiguation (via description)
gpt-5-mini-2025-08-07
Target entity: cuDNN Target entity description: cuDNN is NVIDIA’s GPU-accelerated library of optimized primitives for deep neural networks, widely used to speed up training and inference in frameworks like TensorFlow and PyTorch.
-
A.
NVIDIA CUDA
NVIDIA CUDA is a parallel computing platform and programming model that enables developers to use NVIDIA GPUs for general-purpose high-performance computing.
-
B.
CuPy
CuPy is an open-source array library for Python that accelerates numerical computing by providing a NumPy-compatible interface backed by GPU execution.
-
C.
PlaidML
PlaidML is an open-source, hardware-agnostic deep learning engine designed to accelerate neural network computation on a wide range of GPUs and other devices.
-
D.
NVIDIA AI Enterprise software suite
NVIDIA AI Enterprise software suite is a comprehensive, enterprise-grade collection of AI tools, frameworks, and optimized software designed to accelerate the development and deployment of AI and data analytics workloads across modern data centers and clouds.
-
E.
Microsoft Cognitive Toolkit
Microsoft Cognitive Toolkit (CNTK) is an open-source deep learning framework developed by Microsoft for building, training, and deploying neural networks at scale.
- F. None of above. chosen
Statements (62)
| Predicate | Object |
|---|---|
| instanceOf |
GPU-accelerated library
ⓘ
deep learning library ⓘ |
| acronymFor |
CUDA Deep Neural Network library
ⓘ
linked to:
cuDNN
|
| developer |
NVIDIA
ⓘ
linked to:
NVIDIA Corporation
|
| distribution | NVIDIA Developer website ⓘ |
| domain |
deep learning
ⓘ
neural networks ⓘ |
| fullName |
CUDA Deep Neural Network library
ⓘ
linked to:
cuDNN
|
| goal |
accelerate deep neural network inference
ⓘ
accelerate deep neural network training ⓘ |
| integratesWith |
NVIDIA CUDA ecosystem
ⓘ
NVIDIA TensorRT (indirectly via shared primitives) ⓘ
linked to:
NVIDIA TensorRT
|
| languageBinding |
C
ⓘ
C++ ⓘ |
| license | proprietary ⓘ |
| optimizedFor |
NVIDIA GPU architectures
ⓘ
high throughput ⓘ low latency inference ⓘ |
| programmingModel |
CUDA
ⓘ
linked to:
NVIDIA CUDA
|
| provides |
GPU-accelerated tensor operations
ⓘ
GRU primitives ⓘ LSTM primitives ⓘ activation functions ⓘ high-performance convolution routines ⓘ normalization operations ⓘ optimized primitives for deep neural networks ⓘ pooling operations ⓘ recurrent neural network routines ⓘ tensor layout transformations ⓘ |
| requires |
CUDA Toolkit
ⓘ
linked to:
NVIDIA CUDA
NVIDIA GPU driver ⓘ |
| supportsDataType |
BF16
ⓘ
FP16 ⓘ FP32 ⓘ INT8 ⓘ |
| supportsFeature |
multi-GPU scaling via frameworks
ⓘ
tensor cores acceleration ⓘ |
| supportsHardware |
NVIDIA GPU
ⓘ
linked to:
GPU
NVIDIA data center GPU ⓘ
linked to:
NVIDIA Tesla data center GPUs
NVIDIA embedded GPU ⓘ NVIDIA gaming GPU ⓘ |
| supportsOperation |
RNN inference
ⓘ
RNN training ⓘ backward convolution ⓘ batch normalization ⓘ dropout ⓘ forward convolution ⓘ softmax ⓘ tensor reduction ⓘ |
| targetUser |
deep learning framework developers
ⓘ
high-performance computing developers ⓘ |
| useCase |
inference of deep neural networks
ⓘ
training of deep neural networks ⓘ |
| usedBy |
Caffe
ⓘ
Caffe2 ⓘ Chainer ⓘ MXNet ⓘ Microsoft Cognitive Toolkit ⓘ PaddlePaddle ⓘ PyTorch ⓘ TensorFlow ⓘ Theano ⓘ |
How these facts were elicited
The pipeline generated the facts above by prompting gpt-5.1 with this entity's name + description and the instruction below.
Instruction
You are a knowledge base construction expert. Given a subject entity and a description of it, return factual statements that you know for the subject as a JSON list of dictionaries(triples), where keys must be "subject", "predicate" and "object". The number of facts may be very high, between 25 to 50 or more, for very popular subjects. For less popular subjects, the number of facts can be very low, like 5 or 10. # Requirements - If you don't know the subject at all, return an empty list. - If the subject is not a named entity, return an empty list. - Include at least one triple where predicate is "instanceOf". - Do not get too wordy. - Separate several objects into multiple triples with one object.
Input
Subject: cuDNN Description of subject: cuDNN is NVIDIA’s GPU-accelerated library of optimized primitives for deep neural networks, widely used to speed up training and inference in frameworks like TensorFlow and PyTorch.
Referenced by (24)
Full triples — surface form annotated when it differs from this entity's canonical label.
linked to: cuDNN
subject linked to:
NVIDIA AI Enterprise software suite
linked to: cuDNN
linked to: cuDNN
linked to: cuDNN
subject linked to:
Compute Unified Device Architecture
linked to: cuDNN
subject linked to:
CUDA toolkit
linked to: cuDNN
linked to: cuDNN
linked to: cuDNN