October 23rd, 2026 | 11:30 am
DEIB - NECSTLab Meeting Room (Bld. 20)
Online by Zoom
Contact: Prof. Marco Santambrogio
Abstract
On Friday, October 23rd, 2026, we will have a new talk for the series #NECSTFridayTalk.During this talk, we will have, as speaker, Francesco Pesce, PhD at Dipartimento di Elettronica, Informazione e Bioingegneria.
The rapid growth of sparse machine learning (ML) workloads has highlighted the performance and energy-efficiency limitations of general-purpose architectures, making large-scale deployment increasingly unsustainable economically and environmentally. This has driven the development of domain-specific accelerators (DSAs) for sparse tensor algebra, whose efficiency requires extensive design space exploration (DSE) due to the variability in computation structure, sparsity, and encoding formats.
Efficient DSE remains challenging: general-purpose cycle-accurate simulators such as gem5 are prohibitively slow for realistic datasets, while developing accurate DSA models is complex and error-prone. Specialized tools such as TeAAL and SparseLoop achieve orders-of-magnitude faster simulation while maintaining suitable accuracy on key design metrics, but remain too slow for comprehensive DSE across the diverse real-world tensors required to evaluate sparse tensor architectures.
Turquoise is a hardware generation framework that leverages FPGA acceleration for fast and accurate simulation of sparse algebra DSAs. When modeling the state-of-the-art accelerators Gamma and ExTensor, it achieves accuracy comparable to TeAAL, the leading simulator in this domain, while delivering over two orders of magnitude speed-up on the same dataset. Even including RTL synthesis, implementation, and bitstream generation—steps required only once per architecture—Turquoise achieves an order of magnitude faster simulation, demonstrating the practicality of hardware-assisted DSE for sparse ML DSAs.
The NECSTLab is a DEIB laboratory, with different research lines on advanced topics in computing systems: from architectural characteristics, to hardware-software codesign methodologies, to security and dependability issues of complex system architectures.
#NECSTLab #Computerscience
Every week, the “NECSTFridayTalk” invites researchers, professionals or entrepreneurs to share their work experiences and projects they are implementing in the “Computing Systems”.