6 papers
Interpolation of Non-Linear Functions for LLMs using Partial Reconfiguration in FPGAs
Roger Morales-Monge, Nazareth Jimenez-Chacon, Jose Gabriel Villalobos-Alvarado +2
Non-linear functions such as exponential and sigmoid are essential in AI and LLM acceleration, although implementing them efficiently on FPGAs is still costly. This paper proposes…
CAMTA: A Reconfigurable Multi-Region Activation Unit for Nonlinear Function Approximation
Carlos Soto-Porras, Jose Fonseca-Cruz, Pablo Ramirez-Morera +3
Nonlinear activation functions are widely used in machine learning workloads, but their direct hardware implementation is often costly, function-specific, or difficult to reuse acr…
Design and Implementation of a Multi-Sensor DAQ System for Comparative Photovoltaic Performance Analysis
Maickol Fernandez-Obando, Luis G. Leon-Vega, Leonardo Cardinale-Villalobos +2
The rigorous analysis of specialized physical processes often demands custom data acquisition architectures that offer flexibility and precision beyond the capabilities of general-…
A Quantitative Evaluation of Approximate Softmax Functions for Deep Neural Networks
Anthony Leiva-Valverde, Fabricio Elizondo-Fernández, Luis G. León-Vega +2
The softmax function is a widely used activation function in the output layers of neural networks, responsible for converting raw scores into class probabilities while introducing…
EfiMon: A Process Analyser for Granular Power Consumption Prediction
Luis G. León-Vega, Niccolò Tosato, Stefano Cozzini
High-performance computing (HPC) and supercomputing are critical in Artificial Intelligence (AI) research, development, and deployment. The extensive use of supercomputers for trai…
A Comprehensive Analysis of Process Energy Consumption on Multi-Socket Systems with GPUs
Luis G. León-Vega, Niccolò Tosato, Stefano Cozzini
Robustly estimating energy consumption in High-Performance Computing (HPC) is essential for assessing the energy footprint of modern workloads, particularly in fields such as Artif…