RT Journal Article
T1 Deep PeNSieve: A deep learning framework based on the posit number system
A1 Murillo Montero, Raúl
A1 Del Barrio García, Alberto Antonio
A1 Botella Juan, Guillermo
AB The Posit Number System (PNS) was introduced by John L. Gustafson in 2017. The interesting properties of this novel format can be exploited under the scenario of deep neural networks. In this paper, we propose Deep PeNSieve, a framework for entirely performing both training and inference of deep neural networks employing the PNS. Furthermore, an 8-bit posit quantization approach using fused operations is introduced. In comparison with the state-of-the-art posit frameworks, the proposal has been able to train more complex networks than the feedforward ones, achieving similar accuracies as the floating-point format. The case of CIFAR-10 is especially remarkable, as 16-bit posits even obtain 4% higher top-1 for such dataset. Overall, results show that the proposed quantization approach can preserve model accuracy in the same manner as common quantization techniques.
PB Elsevier
SN 1051-2004
YR 2020
FD 2020-05-07
LK https://hdl.handle.net/20.500.14352/132680
UL https://hdl.handle.net/20.500.14352/132680
LA eng
NO Murillo, R., Del Barrio, A.A. and Botella, G., 2020. Deep PeNSieve: A deep learning framework based on the posit number system. Digital Signal Processing, 102, p.102762.
NO This paper has been supported by the Community of Madrid under grant S2018/TCS-4423, the EU (FEDER) and the Spanish MINECO under grant RTI2018-093684-B-I00 and by Banco Santander under grant PR26/16-20B-1.The source code of the proposed framework, Deep PeNSieve, will be provided as an open-source software and will be available at https://github.com/RaulMurillo/deep-pensieve. This open-source proposal on posit arithmetic would provide a basic platform for more advances and research on posit arithmetic and its application on deep learning.
DS Docta Complutense
RD 20 mar 2026