|
gptkbp:instanceOf
|
AI system
|
|
gptkbp:coreCount
|
2
|
|
gptkbp:countryOfOrigin
|
gptkb:United_States
|
|
gptkbp:dimensions
|
444 mm x 872 mm x 264 mm
|
|
gptkbp:formFactor
|
Rackmount
|
|
gptkbp:GPUArchitecture
|
gptkb:Ampere
|
|
https://www.w3.org/2000/01/rdf-schema#label
|
NVIDIA DGX A100
|
|
gptkbp:manufacturer
|
gptkb:NVIDIA
|
|
gptkbp:maxPerformanceFP16
|
5 petaFLOPS
|
|
gptkbp:maxPerformanceFP64
|
156 teraFLOPS
|
|
gptkbp:maxPerformanceINT8
|
10 petaOPS
|
|
gptkbp:network
|
gptkb:8x_Single-Port_Mellanox_ConnectX-6_VPI_HDR_InfiniBand/Ethernet
|
|
gptkbp:notableUser
|
gptkb:Amazon
gptkb:Argonne_National_Laboratory
gptkb:Baidu
gptkb:Google
gptkb:Lawrence_Livermore_National_Laboratory
gptkb:Microsoft
gptkb:OpenAI
gptkb:Tesla
gptkb:NVIDIA_Research
gptkb:Alibaba
gptkb:Meta
gptkb:Tencent
|
|
gptkbp:numberOfGPUs
|
8
|
|
gptkbp:operatingSystem
|
gptkb:Ubuntu_Linux
|
|
gptkbp:powerSource
|
6,500 W
|
|
gptkbp:predecessor
|
gptkb:NVIDIA_DGX-2
|
|
gptkbp:processor
|
gptkb:NVIDIA_A100_Tensor_Core_GPU
gptkb:AMD_EPYC_7742
|
|
gptkbp:productType
|
gptkb:NVIDIA_DGX
|
|
gptkbp:rackUnits
|
6U
|
|
gptkbp:RAM
|
1 TB
|
|
gptkbp:releaseDate
|
2020
|
|
gptkbp:storage
|
15 TB NVMe SSD
|
|
gptkbp:successor
|
gptkb:NVIDIA_DGX_H100
|
|
gptkbp:targetMarket
|
gptkb:business
data centers
research institutions
|
|
gptkbp:technology
|
gptkb:NVIDIA_TensorRT
gptkb:NVIDIA_cuDNN
gptkb:NVIDIA_GPU_Cloud
gptkb:NVIDIA_CUDA
|
|
gptkbp:totalGPUmemory
|
320 GB
|
|
gptkbp:uses
|
gptkb:machine_learning
AI research
deep learning
|
|
gptkbp:weight
|
134 kg
|
|
gptkbp:bfsParent
|
gptkb:NVLink
gptkb:NVIDIA_DGX_systems
gptkb:DGX
|
|
gptkbp:bfsLayer
|
6
|