Big-PERCIVAL: Exploring the Native Use of 64-Bit Posit Arithmetic in Scientific Computing
Open Access
- 18 March 2024
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computers
- Vol. 73 (6) , 1472-1485
Abstract
The accuracy requirements in many scientific computing workloads result in the use of double-precision floating-point arithmetic in the execution kernels. Nevertheless, emerging real-number representations, such as posit arithmetic, show promise in delivering even higher accuracy in such computations. In this work, we explore the native use of 64-bit posits in a series of numerical benchmarks and compare their timing performance, accuracy and hardware cost to IEEE 754 doubles. In addition, we also study the conjugate gradient method for numerically solving systems of linear equations in real-world applications. For this, we extend the PERCIVAL RISC-V core and the Xposit custom RISC-V extension with posit64 and quire operations. Results show that posit64 can obtain up to 4 orders of magnitude lower mean square error than doubles. This leads to a reduction in the number of iterations required for convergence in some iterative solvers. However, leveraging the quire accumulator register can limit the order of some operations such as matrix multiplications. Furthermore, detailed FPGA and ASIC synthesis results highlight the significant hardware cost of 64-bit posit arithmetic and quire. Despite this, the large accuracy improvements achieved with the same memory bandwidth suggest that posit arithmetic may provide a potential alternative representation for scientific computing.Keywords
All Related Versions
Funding Information
- ERDF A way of making Europe (PID2021-123041OB-I00, PID2021-126576NB-I00)
- CM (S2018/TCS-4423)
This publication has 35 references indexed in Scilit:
- A Suite of Division Algorithms for Posit Arithmetic2023
- PDPU: An Open-Source Posit Dot-Product Unit for Deep Learning Applications2023
- CLARINET: A quire-enabled RISC-V-based framework for posit arithmetic empiricismJournal of Systems Architecture, 2022
- Accelerating Variants of the Conjugate Gradient with the Variable Precision Processor2022
- Comparing Different Decodings for Posit Arithmetic2022
- A Reconfigurable Posit Tensor Unit with Variable-Precision Arithmetic and Automatic Data StreamingJournal of Signal Processing Systems, 2021
- Dynamic Fused Multiply-Accumulate Posit Unit with Variable Exponent Size for Low-Precision DSP Applications2020
- Evaluating the Numerical Stability of Posit Arithmetic2020
- Evaluating the Hardware Cost of the Posit Number System2019
- Parameterized Posit Arithmetic Hardware Generator2018