A 320-MFLOPS CMOS FLOATING-POINT PROCESSING UNIT FOR SUPERSCALAR PROCESSORS

被引:5
作者
IDE, N [1 ]
FUKUHISA, H [1 ]
KONDO, Y [1 ]
YOSHIDA, T [1 ]
NAGAMATSU, M [1 ]
MORI, J [1 ]
YAMAZAKI, I [1 ]
UENO, K [1 ]
机构
[1] TOSHIBA CO LTD,SEMICOND DEVICE ENGN LAB,SAIWAI KU,KAWASAKI 210,JAPAN
关键词
D O I
10.1109/4.210003
中图分类号
TM [电工技术]; TN [电子技术、通信技术];
学科分类号
0808 ; 0809 ;
摘要
A CMOS pipelined floating-point processing unit (FPU) for superscalar processors is described. It is fabricated using a 0.5-mum CMOS triple-metal layer technology on a 61-mm2 die. The FPU has two execution modes to meet precise scientific computations and real-time applications. It can start two FPU operations in each cycle, and this achieves a peak performance of 160 MFLOPS double or single precision with an 80-MHz clock. Furthermore, the original computation mode, twin single-precision computation, doubles the peak performance and delivers 320-MFLOPS single precision. Its full bypass reduces the latency of operations, including load and store, and achieves an effective throughput even in nonvectorizable computations. An out-of-order completion is provided by using a new exception prediction method and a pipeline stall technique.
引用
收藏
页码:352 / 361
页数:10
相关论文
共 23 条
[1]  
AIHARA M, 1990, MAY P CICC
[3]  
BENSCHNEIDER B, 1989, ISSCC FEB, P50
[4]  
BIRMAN M, 1988, OCT P ICCD, P214
[5]  
BOOTH AD, 1951, Q J MECH APPL MATH 2, V4
[6]  
CHAI P, 1991, MAY P CICC
[7]  
DIEFENDORFF K, 1991, NOV P MICR ARCH S JA, P77
[8]  
JOHNSON M, 1991, SUPERSCALAR MICROPRO
[9]  
KIMURA N, 1991, MAY P CICC
[10]  
KOHN L, 1989, FEB ISSCC, P54