A 0.18 μm implementation of a floating-point unit for a processing-in-memory system

Author(s):  
Taek-Jun Kwon ◽  
Joong-Seok Moon ◽  
J. Sondeen ◽  
J. Draper
Author(s):  
Stefan Payer ◽  
Cedric Lichtenau ◽  
Michael Klein ◽  
Kerstin Schelm ◽  
Petra Leber ◽  
...  

Occam is a parallel programming language which can be used to describe VLSI circuits at several levels of abstraction. For a given piece of hardware one might have one Occam description which is close to the implementation level and another which is close to a specification in a notation such as Z or CSP. Thus a design can be substantially verified by a proof of equivalence of such descriptions. This can sometimes be achieved by using transformations based on the algebraic semantics of Occam and, even where this is not possible, the clean design and semantics of Occam make other formal techniques either easier or more reliable. We discuss several case studies: the floating point unit of the IMS T800 transputer and various aspects of its successor, the IMS T9000. Despite the close relationship, on the surface, of these cases studies, radically different techniques were required for them. Based on this and other evidence, the author believes that one of the most important possessions when starting to tackle problems in hardware verification, at least at current levels of knowledge and technology, is an open mind.


Sign in / Sign up

Export Citation Format

Share Document