Statements (49)
Predicate | Object |
---|---|
gptkbp:instanceOf |
supercomputer architecture
|
gptkbp:announced |
2019
|
gptkbp:canBe |
gptkb:DGX-2
gptkb:DGX_A100 gptkb:DGX_H100 |
gptkbp:enables |
gptkb:simulation
data analytics AI model inference AI model training large language model training |
gptkbp:features |
gptkb:NVIDIA_AI_Enterprise
gptkb:NVIDIA_Base_Command reference architecture AI software stack turnkey deployment |
https://www.w3.org/2000/01/rdf-schema#label |
NVIDIA DGX SuperPOD
|
gptkbp:location |
deployed worldwide
|
gptkbp:manufacturer |
gptkb:NVIDIA
|
gptkbp:purpose |
high-performance computing
AI infrastructure |
gptkbp:scalableTo |
thousands of GPUs
|
gptkbp:supports |
machine learning workloads
deep learning workloads |
gptkbp:usedBy |
gptkb:NVIDIA_Cambridge-1
gptkb:Microsoft_Azure NVIDIA Eos NVIDIA Selene |
gptkbp:usedFor |
AI research
climate modeling natural language processing scientific computing speech recognition drug discovery image recognition recommendation systems financial modeling edge AI healthcare AI enterprise AI autonomous vehicle research cloud AI services federated learning data center AI |
gptkbp:uses |
gptkb:NVIDIA_GPUs
gptkb:NVIDIA_DGX_systems gptkb:NVIDIA_NVSwitch NVIDIA InfiniBand networking |
gptkbp:bfsParent |
gptkb:Megatron-Turing_NLG
|
gptkbp:bfsLayer |
6
|