|
gptkbp:instanceOf
|
hardware component
|
|
gptkbp:API
|
gptkb:CUDA
gptkb:TensorFlow
gptkb:TensorRT
gptkb:cuDNN
gptkb:PyTorch
gptkb:DirectML
gptkb:OpenAI_Triton
|
|
gptkbp:architecture
|
systolic array
|
|
gptkbp:developedBy
|
gptkb:NVIDIA
|
|
gptkbp:documentation
|
NVIDIA developer website
|
|
gptkbp:enables
|
gptkb:DLSS_(Deep_Learning_Super_Sampling)
AI denoising
AI upscaling
faster convolution operations
faster deep learning workloads
faster matrix multiplication
real-time ray tracing acceleration
|
|
https://www.w3.org/2000/01/rdf-schema#label
|
Tensor Cores
|
|
gptkbp:integratesWith
|
gptkb:NVIDIA_A100
gptkb:NVIDIA_H100
gptkb:NVIDIA_Quadro_GPUs
gptkb:NVIDIA_Tesla_GPUs
gptkb:NVIDIA_GeForce_GPUs
gptkb:NVIDIA_RTX_30_Series
gptkb:NVIDIA_RTX_40_Series
gptkb:NVIDIA_T4
gptkb:NVIDIA_V100
gptkb:NVIDIA_RTX_GPUs
|
|
gptkbp:introducedIn
|
gptkb:NVIDIA_Volta_architecture
|
|
gptkbp:marketedAs
|
AI acceleration feature
|
|
gptkbp:optimizedFor
|
deep learning
|
|
gptkbp:presentIn
|
gptkb:NVIDIA_Ada_Lovelace
gptkb:NVIDIA_Ampere
gptkb:NVIDIA_Hopper
gptkb:NVIDIA_Turing
gptkb:NVIDIA_Volta
|
|
gptkbp:purpose
|
accelerate matrix operations
|
|
gptkbp:relatedTo
|
GPU acceleration
RT cores
CUDA cores
|
|
gptkbp:supports
|
gptkb:TensorFloat-32
gptkb:BFLOAT16
FP16
INT8
mixed-precision computing
|
|
gptkbp:usedFor
|
high-performance computing
scientific computing
AI inference
AI training
|
|
gptkbp:usedIn
|
gptkb:NVIDIA_GPUs
|
|
gptkbp:bfsParent
|
gptkb:Nvidia_Hopper
gptkb:Nvidia_RTX
gptkb:NVIDIA_GeForce_graphics
|
|
gptkbp:bfsLayer
|
6
|