[HTML][HTML] Design-time methodology for optimizing mixed-precision CPU architectures on FPGA

L Denisov, A Galimberti, D Cattaneo, G Agosta… - Journal of Systems …, 2024 - Elsevier
Approximate computing can significantly reduce the energy consumption of computing
systems. Mixed-precision hardware architectures and precision-tuning tools for software …

Cost-effective fixed-point hardware support for RISC-V embedded systems

D Zoni, A Galimberti - Journal of Systems Architecture, 2022 - Elsevier
With the ever-increasing energy-efficiency requirements for the computing platforms at the
edge, precision tuning techniques highlight the possibility of improving the efficiency of …

Towards extreme scale technologies and accelerators for eurohpc hw/sw supercomputing applications for exascale: the textarossa approach

G Agosta, M Aldinucci, C Alvarez, R Ammendola… - Microprocessors and …, 2022 - Elsevier
In the near future, Exascale systems will need to bridge three technology gaps to achieve
high performance while remaining under tight power constraints: energy efficiency and …

A low-power transprecision floating-point cluster for efficient near-sensor data analytics

F Montagna, S Mach, S Benatti… - … on Parallel and …, 2021 - ieeexplore.ieee.org
Recent applications in low-power (1-20 mW) near-sensor computing require the adoption of
floating-point arithmetic to reconcile high precision results with a wide dynamic range. In this …

Hardware and software support for mixed precision computing: a roadmap for embedded and hpc systems

W Fornaciari, G Agosta, D Cattaneo… - … , Automation & Test …, 2023 - ieeexplore.ieee.org
Mixed precision is an approximate computing technique that can be used to trade-off
computation accuracy for performance and/or energy. It can be applied to many error …

Adaptive R-peak detection on wearable ECG sensors for high-intensity exercise

E De Giovanni, T Teijeiro, GP Millet… - IEEE Transactions on …, 2022 - ieeexplore.ieee.org
Objective: Continuous monitoring of biosignals via wearable sensors has quickly expanded
in the medical and wellness fields. At rest, automatic detection of vital parameters is …

Optimization of the fixed-point representation of measurement data for intelligent measurement systems

MR Dinčić, ZH Perić, DB Denić, BD Denić - Measurement, 2023 - Elsevier
The development of intelligent measurement systems by implementing AI (artificial
intelligence) algorithms on edge sensor nodes is very topical. Due to very limited hardware …

An efficient multi-format low-precision floating-point multiplier

HA Kermani, AAE Zarandi - Sustainable Computing: Informatics and …, 2024 - Elsevier
Low-precision computing has emerged as a promising technology to enhance performance
in modern applications like deep neural network training and scientific computing. However …

Textarossa: Towards extreme scale technologies and accelerators for eurohpc hw/sw supercomputing applications for exascale

G Agosta, D Cattaneo, W Fornaciari… - 2021 24th Euromicro …, 2021 - ieeexplore.ieee.org
To achieve high performance and high energy efficiency on near-future exascale computing
systems, three key technology gaps needs to be bridged. These gaps include: energy …

Analysis of 32-bit Fixed Point Quantizer in the Wide Variance Range for the Laplacian Source

Z Perić, A Jovanović, M Dinčić, M Savić… - … and Services in …, 2021 - ieeexplore.ieee.org
The main goal of this paper is to examine the possibility of using the 32-bit fixed-point format
to represent the weights of neural networks (NN) instead of the standardly used 32-bit …