Constraint-aware evolutionary algorithms for auto-tuning improve efficiency by ~39x on average by skipping invalid hardware/software configurations, outperforming pyATF on benchmarks.
Fftw: An adaptive software architecture for the fft,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.NE 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Constraint-aware Optimization in Auto-Tuning
Constraint-aware evolutionary algorithms for auto-tuning improve efficiency by ~39x on average by skipping invalid hardware/software configurations, outperforming pyATF on benchmarks.