Filters








18 Hits in 2.8 sec

The transprecision computing paradigm: Concept, design, and applications

A. Cristiano I. Malossi, Michael Schaffner, Anca Molnos, Luca Gammaitoni, Giuseppe Tagliavini, Andrew Emerson, Andres Tomas, Dimitrios S. Nikolopoulos, Eric Flamand, Norbert Wehn
2018 2018 Design, Automation & Test in Europe Conference & Exhibition (DATE)  
To overcome the power wall, a shift from traditional computing paradigms is now mandatory.  ...  Indeed, OPRECOMP aims at demolishing the ultraconservative "precise" computing abstraction, replacing it with a more flexible and efficient one, namely transprecision computing.  ...  This paper is organized as follows: in Section II we introduce and define the transprecision computing paradigm.  ... 
doi:10.23919/date.2018.8342176 dblp:conf/date/MalossiSMGTETNF18 fatcat:kjjsekbtdrd3lbgjkhf6ki27ne

A Transprecision Floating-Point Architecture for Energy-Efficient Embedded Computing

Stefan Mach, Davide Rossi, Giuseppe Tagliavini, Andrea Marongiu, Luca Benini
2018 2018 IEEE International Symposium on Circuits and Systems (ISCAS)  
In light of transprecision computing, where accuracy of data is consciously changed during the execution of applications, custom FP types are being used to optimize a wide range of problems.  ...  The rising computational demands and high dynamic of target algorithms often call for hardware support of floating-point (FP) arithmetic and high system energy efficiency.  ...  ACKNOWLEDGEMENTS This project has received funding from the European Union's Horizon 2020 research and innovation programme under grant agreement No 732631 (OPRECOMP).  ... 
doi:10.1109/iscas.2018.8351816 dblp:conf/iscas/MachRTMB18 fatcat:kao36jyua5cj7cwtyvmbo6vsay

A transprecision floating-point platform for ultra-low power computing

Giuseppe Tagliavini, Stefan Mach, Davide Rossi, Andrea Marongiu, Luca Benin
2018 2018 Design, Automation & Test in Europe Conference & Exhibition (DATE)  
Finally, we present the design of a transprecision FP unit capable of handling 8-bit and 16-bit operations in addition to standard 32bit operations.  ...  From a theoretical point of view, the adoption of multiple FP types perfectly fits with the principle of transprecision computing, allowing fine-grained control of approximation while meeting specified  ...  ACKNOWLEDGMENT This work has been partially supported by the European FP7 ERC project MULTITHERMAN (g.a. 291125) and by the European H2020 FET project OPRECOMP (g.a. 732631).  ... 
doi:10.23919/date.2018.8342167 dblp:conf/date/TagliaviniMRMB18 fatcat:cjfsxpi2sjgwznj4h7rezjpeai

A Transprecision Floating-Point Platform for Ultra-Low Power Computing [article]

Giuseppe Tagliavini, Stefan Mach, Davide Rossi, Andrea Marongiu, Luca Benini
2017 arXiv   pre-print
Finally, we present the design of a transprecision FP unit capable of handling 8-bit and 16-bit operations in addition to standard 32-bit operations.  ...  In modern low-power embedded platforms, floating-point (FP) operations emerge as a major contributor to the energy consumption of compute-intensive applications with large dynamic range.  ...  ACKNOWLEDGMENT This work has been partially supported by the European FP7 ERC project MULTITHERMAN (g.a. 291125) and by the European H2020 FET project OPRECOMP (g.a. 732631).  ... 
arXiv:1711.10374v1 fatcat:roeu3zx6nnahtb5measfypdm4e

Design and Evaluation of SmallFloat SIMD extensions to the RISC-V ISA

Giuseppe Tagliavini, Stefan Mach, Davide Rossi, Andrea Marongiu, Luca Benini
2019 2019 Design, Automation & Test in Europe Conference & Exhibition (DATE)  
Motivated by the important energy savings that smaller-than-32-bit FP types have enabled in several application domains and related compute platforms, some recent studies have published encouraging early  ...  The RISC-V 32IMFC ISA configuration has been widely adopted for the design of newgeneration, low-power processors.  ...  The most promising approaches are moving beyond the concept of approximation alone, toward a novel paradigm called transprecision computing [12] , which aims at designing system to deliver just the required  ... 
doi:10.23919/date.2019.8714897 dblp:conf/date/TagliaviniMRMB19 fatcat:sf7v4c526zakzpeaswml6h25hi

FlexFloat: A Software Library for Transprecision Computing

Giuseppe Tagliavini, Andrea Marongiu, Luca Benini
2018 IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems  
In recent years approximate computing has been extensively explored as a paradigm to design hardware and software solutions that save energy by trading off on the quality of the computed results.  ...  Its design allows to emulate the behavior of standard IEEE FP types and custom extensions for reduced-precision computation.  ...  Abstract-In recent years approximate computing has been extensively explored as a paradigm to design hardware and software solutions that save energy by trading off on the quality of the computed results  ... 
doi:10.1109/tcad.2018.2883902 fatcat:ya3uyzryfzbbbl42bksrcgielq

Agile Autotuning of a Transprecision Tensor Accelerator Overlay for TVM Compiler Stack [article]

Dionysios Diamantopoulos, Burkhard Ringlein, Mitra Purandare, Gagandeep Singh, Christoph Hagleitner
2020 arXiv   pre-print
computing.  ...  The rapid development of frameworks, models, and precision options challenges the adaptability of such tensor-accelerators since the adaptation to new requirements incurs significant engineering costs.  ...  Recently, transprecision computing was proposed as a paradigm shift in precision selection, and suggests the adaptability of precision according to an application's requirements [20] , as opposed to the  ... 
arXiv:2004.10854v1 fatcat:i4gkoxo2nbdf5l3wo5cln77lsy

Guest Editorial for the Special Issue on Energy-Quality Scalable Circuits and Systems for Sensing and Computing: from Approximate, to Communication-Inspired and Learning-Based

Massimo Alioto, Vivek De, Andrea Marongiu
2018 IEEE Journal on Emerging and Selected Topics in Circuits and Systems  
S USTAINED energy efficiency improvements have been instrumental for the vertiginous evolution of electronic systems with computing, sensing, communication and storage capabilities.  ...  Energy reductions are also critical in portable electronics, due to the limited thermal budget and battery energy availability.  ...  Eduard Alarcon for his unconditioned support and valuable insights, and the JETCAS admin Ms Desiree Noel for her precious help with managing the submissions and the review process.  ... 
doi:10.1109/jetcas.2018.2865783 fatcat:w7pvflbf2rfxraswc3hqijzr44

Toward a modular precision ecosystem for high-performance computing

Hartwig Anzt, Goran Flegar, Thomas Grützmacher, Enrique S Quintana-Ortí
2019 The international journal of high performance computing applications  
Concretely, the goal is to radically decouple the data storage format from the processing format and, ultimately, design a "modular precision ecosystem" that allows for more flexibility in terms of customized  ...  With the memory bandwidth of current computer architectures being significantly slower than the (floating point) arithmetic performance, many scientific computations only leverage a fraction of the computational  ...  Quintana-Ortí were supported by project TIN2017-82972-R of the MINECO and FEDER and the H2020 EU FETHPC Project 732631 "OPRECOMP".  ... 
doi:10.1177/1094342019846547 fatcat:63sikoiq5nfzzo2rb5wkidzdwe

A Transprecision Floating-Point Architecture for Energy-Efficient Embedded Computing

Stefan Mach, Davide Rossi, Giuseppe Tagliavini, Andrea Marongiu, Luca Benini
2018
In light of transprecision computing, where accuracy of data is consciously changed during the execution of applications, custom FP types are being used to optimize a wide range of problems.  ...  The rising computational demands and high dynamic of target algorithms often call for hardware support of floating-point (FP) arithmetic and high system energy efficiency.  ...  ACKNOWLEDGEMENTS This project has received funding from the European Union's Horizon 2020 research and innovation programme under grant agreement No 732631 (OPRECOMP).  ... 
doi:10.3929/ethz-b-000310292 fatcat:4qhhwrisvzhp7b74txnrg6c2tq

2021 Index IEEE Transactions on Very Large Scale Integration (VLSI) Systems Vol. 29

2021 IEEE Transactions on Very Large Scale Integration (vlsi) Systems  
The primary entry includes the coauthors' names, the title of the paper or other item, and its location, specified by the publication abbreviation, year, month, and inclusive pagination.  ...  The Subject Index contains entries describing the item under all appropriate subject headings, plus the first author's name, the publication abbreviation, month, and year, and inclusive pages.  ...  1782-1789 Azar, K., +, TVLSI April 2021 643-656 Bisheh-Niasar, M., +, TVLSI July 2021 1297-1305 Energy-Proportional Transprecision Computing.  ... 
doi:10.1109/tvlsi.2021.3136367 fatcat:fwqswbyzejgfhgbzywrvsf2qgi

Self-Adaptive Run-Time Variable Floating-Point Precision for Iterative Algorithms: A Joint HW/SW Approach

Noureddine Ait Said, Mounir Benabdenbi, Katell Morin-Allory
2021 Electronics  
Both algorithms are analyzed and compared statistically to understand the effects of VPT on iterative applications.  ...  In this work, we propose a new method and an FPU architecture that enable designers to dynamically tune FP computations' precision automatically at run-time called Variable Precision in Time (VPT), leading  ...  Transprecision Computing (TC) [1] is a paradigm that came to implement this vision by introducing efficient, flexible, and reconfigurable architectures adequate for such applications.  ... 
doi:10.3390/electronics10182209 fatcat:ufft3u3hnnh6hjhxc34xthhsre

Dynamic Precision Autotuning with TAFFO

Stefano Cherubin, Daniele Cattaneo, Michele Chiari, Giovanni Agosta
2020 ACM Transactions on Architecture and Code Optimization (TACO)  
Many classes of applications, both in the embedded and high performance domains, can trade off the accuracy of the computed results for computation performance.  ...  Within approximate computing, one specific technique, precision tuning, deals with leveraging the trade off between the error introduced in the computation and the performance of the application by changing  ...  Static mixed precision tuning is performed once for every application and it is considered part of the system design.  ... 
doi:10.1145/3388785 fatcat:a5crxzmlzfdoded7y4nq6dlv4u

Analysis of Threading Libraries for High Performance Computing

Adrian Castello, Rafael Mayo, Sangmin Seo, Pavan Balaji, Enrique S. Quintana-Orti, Antonio J. Pena
2020 IEEE transactions on computers  
With the appearance of multi-/many core machines, applications and runtime systems evolved in order to exploit the new on-node concurrency brought by new software paradigms.  ...  Moreover, we study the semantics offered by threading libraries in order to expose the similarities among different LWT application programming interfaces and their advantages over Pthreads.  ...  Acknowledgments The researchers from the Universitat Jaume I and Uni-  ... 
doi:10.1109/tc.2020.2970706 fatcat:3yeebkavkbh5dj36hr4feqochm

An Open-Source Platform for High-Performance Non-Coherent On-Chip Communication [article]

Andreas Kurth, Wolfgang Rönninger, Thomas Benz, Matheus Cavalcante, Fabian Schuiki, Florian Zaruba, Luca Benini
2020 arXiv   pre-print
We design and implement a state-of-the-art ML training accelerator, where our communication fabric scales to 1024 cores on a die, providing 32 TB/s cross-sectional bandwidth at only 24 ns round-trip latency  ...  On-chip communication infrastructure is a central component of modern systems-on-chip (SoCs), and it continues to gain importance as the number of cores, the heterogeneity of components, and the on-chip  ...  His research interests include computer architecture, transprecision computing, as well as near-memory and in-memory processing.  ... 
arXiv:2009.05334v1 fatcat:42ttl7lytzcx3npvc7gjvtabjy
« Previous Showing results 1 — 15 out of 18 results