Presentation
Tensor Cores for Matrix Multiplication Are on the Rise – Is There Any Hope for Sparse Operations?
DescriptionIn response to the demands from the machine learning community, an increasing number of hardware architectures feature tensor cores for high performance dense matrix multiplication in low precision. This is increasing the peak performance and power draw - and the expectation that applications achieve higher performance. But at the same time, many application sparse linear algebra is generally unable to exploit the high performance tensor cores. Is there any hope for improved performance in applications based on sparse linear algebra?
