1 paper
Robert Hochgraf, Sreepathi Pai
Hand-optimizing linear algebra kernels for different GPU devices and applications is complex and labor-intensive. Instead, many developers use automatic performance tuning (autotun…