A5062995565- Parallel Computing and Optimization Techniques
- Advanced Data Storage Technologies
- Matrix Theory and Algorithms
- Face and Expression Recognition
- Algorithms and Data Compression
- Low-power high-performance VLSI design
- Tensor decomposition and applications
- Error Correcting Code Techniques
- Advanced Clustering Algorithms Research
- Numerical Methods and Algorithms
University of Trento
2025
Cornell University
2024-2025
Lawrence Berkeley National Laboratory
2023
The optimization of the matrix multiplication (or GEMM) has been a need during last decades. This operation is considered flagship current linear algebra libraries such as BLIS, OpenBLAS, or Intel OneAPI because its widespread use in large variety scientific applications. GEMM usually implemented following GotoBLAS philosophy, which tiles operands and uses series nested loops for performance improvement. These approaches extract maximum computational power architectures through small pieces...
K-means is a popular clustering algorithm with significant applications in numerous scientific and engineering areas. One drawback of its inability to identify non-linearly separable clusters, which may lead inaccurate solutions certain cases. Kernel variant classical that can find clusters. However, it scales quadratically respect the size dataset, taking several minutes cluster even medium-sized datasets on traditional CPU-based machines. In this paper, we present formulation using...
Sparse symmetric positive definite systems of equations are ubiquitous in scientific workloads and applications. Parallel sparse Cholesky factorization is the method choice for solving such linear systems. Therefore, development parallel codes that can efficiently run on today's large-scale heterogeneous distributed-memory platforms vital importance. Modern supercomputers offer nodes contain a mix CPUs GPUs. To fully utilize computing power these nodes, must be adapted to offload expensive...
The optimization of the matrix multiplication (or GEMM) has been a need during last decades. This operation is considered flagship current linear algebra libraries such as BLIS, OpenBLAS, or Intel OneAPI because its widespread use in large variety scientific applications. GEMM usually implemented following GotoBLAS philosophy, which tiles operands and uses series nested loops for performance improvement. These approaches extract maximum computational power architectures through small pieces...