Re: Unix spreadsheet in SpecInt92

In article <1993Nov4.070829.3291_at_leland.Stanford.EDU> bolguin_at_leland.Stanford.EDU (Ben Olguin) writes:
>If I remember correctly, the SpecInt92 uses Unix spreadsheet "sc"
>program. Are values with 2 decimal places not considered as
>needing to use a FPU in a spreadsheet, even if I have to do all
>kinds of math operations on them? Is the case that "floating
>point" operations does not refer to decimal point values that
>might crop up in accounting operations performed on
>spreadsheets? Or is the requirement that I do not use values with
>a decimal point in accunting spreadsheets, but only round integer
>numbers?
>
>I'd appreciate an explanation from this community. :)
>
I am not quite sure what you are asking but I'll give it a try... :-)

In the creation of the SPEC CINT92 and CFP92 benchmarks suites, it was determined that a program would be placed in CFP92, the floating point suite, if more than 1.0% of the total instructions executed were floating point instructions. If there were less than 1.0% then it was a candidate for CINT92, the non-floating point or integer suite.

The SPEC benchmark 072.sc performs a series of spreadsheet operations. Among the tasks used in the SPEC input set are the creation of a balance sheet, a mortgage calculation and some SPEC metric calculations (geometric means). All of these are floating point operations. However, all of these floating point operations account for less than 1% of the instructions executed (as reported by the various SPEC members on various hardware platforms during benchmark development).

Jeff
Release Manager for CINT92/CFP92

Jeff Reilly			| "There is something fascinating about
Intel Corporation		| science. One gets such wholesale returns
jwreilly_at_mipos2.intel.com	| of conjecture out of such a trifling
(408) 765 - 5909		| investment of fact" - M. Twain