Statements (25)
| Predicate | Object |
|---|---|
| gptkbp:instanceOf |
gptkb:technology
|
| gptkbp:category |
data transfer technology
GPU computing |
| gptkbp:compatibleWith |
gptkb:NVIDIA_GPUs
NVMe storage InfiniBand networks |
| gptkbp:developedBy |
gptkb:NVIDIA
|
| gptkbp:enables |
direct memory access between GPUs and third-party devices
|
| gptkbp:firstReleased |
2010
|
| gptkbp:hasComponent |
GPUDirect Peer-to-Peer
GPUDirect RDMA GPUDirect Storage |
| https://www.w3.org/2000/01/rdf-schema#label |
NVIDIA GPUDirect
|
| gptkbp:improves |
data transfer speed
|
| gptkbp:reduces |
CPU overhead in data transfers
|
| gptkbp:supports |
RDMA (Remote Direct Memory Access)
|
| gptkbp:usedFor |
high-performance computing
data transfer optimization |
| gptkbp:usedIn |
gptkb:machine_learning
data centers AI workloads |
| gptkbp:bfsParent |
gptkb:Nvidia_A100_GPU
gptkb:Quadro_K_series gptkb:Tesla_P100 |
| gptkbp:bfsLayer |
7
|