High-Level Synthesis Optimization for Blocked Floating-Point Matrix Multiplication
نویسندگان
چکیده
منابع مشابه
FPGA accelerator for floating-point matrix multiplication
This study treats architecture and implementation of a FPGA accelerator for double-precision floating-point matrix multiplication. The architecture is oriented towards minimising resource utilisation and maximising clock frequency. It employs the block matrix multiplication algorithm which returns the result blocks to the host processor as soon as they are computed. This avoids output buffering...
متن کاملScalable Floating-Point Matrix Inversion Design Using Vivado High-Level Synthesis (XAPP1317)
Matrix inversion is widely used in numerous signal processing and data analysis algorithms. Many of these algorithms use a floating-point data format to accommodate large dynamic ranges of the random matrices (see Adaptive Beamforming for Radar: Floating-Point QRD+WBS in an FPGA (WP452) [Ref 1]). This application note focuses on the design of a scalable matrix inversion function using the Vivad...
متن کاملEnergy Performance of Floating-Point Matrix Multiplication on FPGAs
Floating-point matrix multiplication is a basic kernel in scientific computing. It has been shown that implementations of this kernel on FPGAs can achieve high sustained performance [1]. However, to the best of our knowledge, existing work on FPGA-based floating-point matrix multiplication considers the optimization of latency or area only. In this paper, we analyze the impact of various parame...
متن کاملError bounds on complex floating-point multiplication
Given floating-point arithmetic with t-digit base-β significands in which all arithmetic operations are performed as if calculated to infinite precision and rounded to a nearest representable value, we prove that the product of complex values z0 and z1 can be computed with maximum absolute error |z0| |z1| 12β √ 5. In particular, this provides relative error bounds of 2−24 √ 5 and 2−53 √ 5 for I...
متن کاملEfficient Hybrid Method for Binary Floating Point Multiplication
This paper presents a high speed binary floating point multiplier based on Hybrid Method. To improve speed multiplication of mantissa is done using Hybrid method replacing existing multipliers like Carry Save Multiplier, Dadda Multiplier and Modified Booth Multiplier. Hybrid method is a combination of Dadda Multiplier and Modified Radix-8 Booth Multiplier. The design achieves high speed with ma...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: ACM SIGARCH Computer Architecture News
سال: 2017
ISSN: 0163-5964
DOI: 10.1145/3039902.3039916