首页> 外文会议>IEEE International Conference on Smart Cloud >Addressing Inefficiency of Floating-Point Operations in Cloud Computing: Implementation and a Case Study of Variable Precision Computing
【24h】

Addressing Inefficiency of Floating-Point Operations in Cloud Computing: Implementation and a Case Study of Variable Precision Computing

机译:解决云计算中浮点运算的效率低下:可变精度计算的实现和案例研究

获取原文

摘要

Cloud computing is increasingly becoming a significant solution for providing scalable computing resources via Internet [1]. However, as the increasing of data size in cloud computing applications and precision requirement in some special scenarios, bandwidth and precision are highly valued in floating-point associated applications, optimizing existing floating-point arithmetic is a significant way to improve current situation. Unum (universal number) arithmetic is a new floating-point arithmetic presented by John L. Gustafson in 2013, being compared with IEEE 754 floats, the outstanding features of unum are clearance of rounding errors, high informationper-bit and variable precision. However, unum was only implemented in software before due to technical complexity, we implement unum arithmetic on FPGA for the first time. We also implement unum based 16-point FFT on FPGA to demonstrate the advantages of unum arithmetic against IEEE 754 floats. We validate the design and contrast the precision and bit width in computing with IEEE 754 floats, evaluate the power efficiency of the designed module. The comparison results show that unum need less bit width than IEEE 754 floats under same precision, it can decrease bandwidth requirement in computing, and ensure correctness even in some extreme arithmetic cases in which IEEE 754 floats cannot work properly.
机译:云计算越来越多地成为通过Internet提供可扩展计算资源的重要解决方案[1]。然而,由于云计算应用中的数据大小的增加和某些特殊情景中的精度要求,带宽和精度在浮点相关应用中受到高度重视,因此优化现有的浮点算法是提高当前情况的重要方法。 UNUM(通用数字)算术是John L. Gustafson于2013年呈现的新浮点算术,与IEEE 754浮动相比,UNUM的出色特征是舍入误差,高信息检测器和可变精度的清除。但是,UNUM在技术复杂性之前仅在软件中实施,我们第一次在FPGA上实施UNUM算术。我们还在FPGA上实施了基于UNUM的16点FFT,以展示UNUM算法对IEEE 754浮子的优势。我们验证了设计和对比使用IEEE 754浮点数计算的精度和比特宽度,评估设计模块的功率效率。比较结果表明,UNUM需要比IEEE 754漂浮的比特宽度相同,它可以降低计算中的带宽要求,并确保即使在某些极端算术情况下,IEEE 754浮动不能正常工作。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号