TECHNOLOGY PROFILE · TECH-0043
cuDNN
NVIDIA library of neural-network computation primitives.
Reference link only — not reviewed this edition. This profile preserves the prepared catalog’s original descriptions. Advantages depend on the workload; comparisons are editorial, not benchmark results.
Where it has an edge
Optimized implementations and execution plans avoid hand-writing common GPU neural operators.
Best-fit problems
CNN and neural-kernel workloads in the NVIDIA ecosystem.
Tradeoffs & cautions
Supported operations and performance depend on device, shape and precision.
What to test before choosing
Benchmark exact shapes, dtypes and devices after warm-up; verify numerical tolerances and end-to-end speed.
Runtime & hardware
Hardware-specific; see profile
Access & licensing
Mixed open/proprietary components; verify terms
Execution
Local / self-hosted
Lifecycle
Not independently audited
Comparison basis
Editorial problem-fit judgement; no head-to-head benchmark performed
Connected concepts
Alternatives to investigate
CUTLASS · oneDNN