05 Feb Chronos is GPU ready
The paper with the results of the multi-GPU implementation of Chronos is available here and it has been submitted to IJHPCA.
In the paper we show how the adaptive FSAI, an approximate inverse characterized by a very high degree of parallelism, can be successfully implemented on a distributed memory computer equipped with GPU accelerators. Taking advantage of GPUs in adaptive FSAI set-up is not a trivial task, nevertheless we show through an extensive numerical experimentation how the proposed approach outperforms more traditional preconditioners and results in a close-to-ideal behaviour in challenging linear algebra problems.
Next step is the porting on the FPGA systems.
The Chronos library is available for research purpose at http://www.m3eweb.it/chronos/
Marconi 100 Supercomputer (courtesy of CINECA)