arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2308.03425cs.ARcs.PF
收藏

PPU:一种流水线式完整 Posit 处理单元的设计与实现

PPU: Design and Implementation of a Pipelined Full Posit Processing Unit

Federico Rossi, Francesco Urbani, Marco Cococcioni, Emanuele Ruffaldi, Sergio Saponara

AI总结:

本文设计并实现了一种流水线式完整 Posit 处理单元 FPPU,将其集成到低功耗 Ibex RISC-V 核心中,以较小面积开销提供实数处理能力,并在深度神经网络中验证了 16 位 posit 可替代 32 位 IEEE 浮点数且精度损失极小。

摘要:

通过利用模块化 RISC-V ISA,本文提出了使用 posit™ 算术指令定制指令集,以提供改进的数值精度、明确定义的行为和更大的可表示数范围,同时保持开源 ISA 的灵活性和优势,例如无需许可和版税费用以及社区开发。在这项工作中,我们介绍了一个完整 posit 处理单元的设计、实现及其在低功耗 Ibex RISC-V 核心中的集成,该单元能够直接在硬件中实现四种算术运算(加、减、乘、除和 fma)、取反、float 到 posit 以及 posit 到 float 的转换。我们评估了该单元(我们称之为 Full Posit Processing Unit,FPPU)的速度、功耗和面积。FPPU 已在 Alveo 和 Kintex FPGA 上进行了原型验证,并评估了其对完整 RISC-V 核心指标的影响,结果表明,我们可以为所述核心提供实数处理能力,8 位 posit 的面积增加限制在 7%,16 位 posit 的面积增加限制在 15%。最后,我们展示了在不同网络模型和数据集上将 posit 用于深度神经网络的测试,结果表明,使用 16 位 posit 代替 32 位 IEEE 浮点数时精度下降极小。

英文摘要:

By exploiting the modular RISC-V ISA this paper presents the customization of instruction set with posit\textsuperscript{\texttrademark} arithmetic instructions to provide improved numerical accuracy, well-defined behavior and increased range of representable numbers while keeping the flexibility and benefits of open-source ISA, like no licensing and royalty fee and community development. In this work we present the design, implementation and integration into the low-power Ibex RISC-V core of a full posit processing unit capable to directly implement in hardware the four arithmetic operations (add, sub, mul, div and fma), the inversion, the float-to-posit and posit-to-float conversions. We evaluate speed, power and area of this unit (that we have called Full Posit Processing Unit). The FPPU has been prototyped on Alveo and Kintex FPGAs, and its impact on the metrics of the full-RISC-V core have been evaluated, showing that we can provide real number processing capabilities to the mentioned core with an increase in area limited to $7\%$ for 8-bit posits and to $15\%$ for 16-bit posits. Finally we present tests one the use of posits for deep neural networks with different network models and datasets, showing minimal drop in accuracy when using 16-bit posits instead of 32-bit IEEE floats.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑