A new VLSI vector arithmetic coprocessor for the PC

A new vector arithmetic coprocessor MIM XPA3233 with integrated PCI bus interface has been developed in CMOS VLSI technology. The chip performs dot products of vectors with components of the IEEE DOUBLE data format to full accuracy or with only one final rounding. Details on the realisation of the multiplication, accumulation and carry resolution processes are discussed. Performance data and some details about the actual VLSI realisation are presented. Software support for the coprocessor is available in the programming languages PASCAL-XSC and C-XSC or from a special C subroutine library. Programming examples are shown using PASCAL-XSC and C.<<ETX>>

[1]  A. Knofel,et al.  Fast hardware units for the computation of accurate dot products , 1991, [1991] Proceedings 10th IEEE Symposium on Computer Arithmetic.

[2]  Ulrich W. Kulisch,et al.  Numerical Toolbox for Verified Computing I , 1993 .

[3]  Ulrich W. Kulisch,et al.  PASCAL-XSC , 1992, Springer Berlin Heidelberg.

[4]  Willard L. Miranker,et al.  A new approach to scientific computation , 1983 .

[5]  G. Bohlender.  What Do We Need Beyond IEEE Arithmetic? , 1990, Computer Arithmetic and Self-Validating Numerical Methods.

[6]  Andreas Knöfel.  Hardwareentwurf eines Rechenwerkes für semimorphe Skalar- und Vektoroperationen unter Berücksichtigung der Anforderungen verifizierender Algorithmen , 1991 .

[7]  B. Hoefflinger,et al.  The CMOS gate forest: an efficient and flexible high-performance ASIC design environment , 1988, IEEE J. Solid State Circuits.

[8]  Wolfgang Rülling,et al.  Exact accumulation of floating-point numbers , 1991, IEEE Symposium on Computer Arithmetic.

[9]  Christoph Baumhof,et al.  A CMOS Floating-Point Processing Chip for Verified Exact Vector Arithmetic , 1994, ESSCIRC '94: Twientieth European Solid-State Circuits Conference.