Fabrizio Petrini is a Principal Engineer of the Intel Parallel Computing Labs in Santa Clara, CA. He is one of industry's lead researchers driving the convergence of High Performance Computing and High Performance Data Analytics, with particular focus on large-scale real-time graph analytics, sparse linear algebra, and unsupervised learning based on tensor algorithms. His research interests cover the spectrum of parallel algorithms, the architectural implications of these algorithms, the design of future architectures to efficiently support them, and the transition of these concepts into viable commercial systems. He is the Co-PI of PUMA, Intel's response to DARPA HIVE.
Accelerators, FPGA, and GPUs
Reliability and Resiliency
System Software and Runtime Systems