Hardware support for multivariable floating point function generation

A method suited for hardware implementation assists in evaluating a multiva riable, floating-point function F(X,Y,...Z) defined on an N-dimensional rec tangular grid. A preprocessing stage analyzes the function's breakpoints fo r each dimension. If determines a subinterval range that includes the break point range for flint dimension and subdivides the subinterval range into e qual-sized subintervals in a procedure that enables easy identification of the breakpoint interval containing a subinterval. The method's run-time sta ge processes the N components of the input argument in parallel to identify the N-dimensional rectangular cell containing the input argument. At run t ime, two additions and a multiplication convert mch component of the input argument to ail index identifying the subinterval containing it. Then two t able lookups identify the breakpoint interval containing the sub-interval. An implementation of a two-variable function generator that uses bilinear i nterpolation is sketched lo illustrate application of the cell-identificati on method.