Nvidia H100 Tensor Core GPU

GPTKB entity

Predicate	Object
gptkbp:instanceOf	gptkb:graphics_card
gptkbp:application	Data analytics AI inference AI training High Performance Computing
gptkbp:architecture	gptkb:Hopper
gptkbp:coreCount	528 16896
gptkbp:formFactor	PCIe card SXM5 module
gptkbp:interface	gptkb:NVLink gptkb:PCIe_5.0
gptkbp:manufacturer	gptkb:Nvidia
gptkbp:memoryBusWidth	3 TB/s
gptkbp:memoryType	gptkb:HBM3
gptkbp:peakFP64Performance	1,979 TFLOPS 67 TFLOPS 34 TFLOPS
gptkbp:peakTensorPerformance	3,958 TFLOPS
gptkbp:powerSource	350W 700W
gptkbp:processNode	gptkb:TSMC_4N
gptkbp:RAM	80 GB 94 GB
gptkbp:releaseDate	2022
gptkbp:successor	gptkb:Nvidia_A100_Tensor_Core_GPU
gptkbp:supports	gptkb:NVLink gptkb:Confidential_Computing gptkb:Multi-Instance_GPU_(MIG) gptkb:PCIe_5.0 gptkb:Transformer_Engine gptkb:NVSwitch gptkb:DPX_Instructions gptkb:Secure_Boot BF16 FP16 INT8 FP32 FP64 ECC Memory TF32 Sparsity Acceleration
gptkbp:transistorCount	80 billion
gptkbp:usedIn	gptkb:Nvidia_DGX_H100 gptkb:Nvidia_HGX_H100 cloud computing supercomputers
gptkbp:bfsParent	gptkb:Nvidia_HGX
gptkbp:bfsLayer	7
https://www.w3.org/2000/01/rdf-schema#label	Nvidia H100 Tensor Core GPU