Confidential GPU inference under model swapping is significantly slower than non-confidential, and the gap is driven by model loading, not by inference compute.
Nvidia h100 tensor core gpu architecture,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.PF 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
Performance of Confidential Computing GPUs
Confidential GPU inference under model swapping is significantly slower than non-confidential, and the gap is driven by model loading, not by inference compute.