home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: sci.math.num-analysis
- Path: sparky!uunet!europa.asd.contel.com!darwin.sura.net!zaphod.mps.ohio-state.edu!moe.ksu.ksu.edu!ux1.cso.uiuc.edu!csrd.uiuc.edu!news
- From: hblim@sp1.csrd.uiuc.edu (Hock-Beng Lim)
- Subject: 32-bit vs 64-bit precision ?
- Message-ID: <1992Nov17.115111.23520@csrd.uiuc.edu>
- Sender: news@csrd.uiuc.edu
- Organization: Univ of Illinois, Center for Supercomputing R&D, Urbana, IL
- Date: Tue, 17 Nov 92 11:51:11 GMT
- Lines: 19
-
- I have a question which I hope the numerical analysis gurus can help
- me out. I would like to find out what are the factors to consider
- when deciding whether to implement a numerical algorithm in 32-bit
- or 64-bit floating-point precision, apart from the obvious difference
- in their precision.
-
- For example, are there any numerical properties which might make an
- algorithm stable when implemented in 64-bit precision ? I heard that Monte
- Carlo techniques are more stable when implemented in 64-bit precision. Is
- that true ? What about some other such differences for other well-known
- algorithms ? The algorithms I am primarily interested in are the linear
- system solvers and the eigensolvers.
-
- HB
- --
- ---------------------------------------------------------------------
- Hock-Beng Lim Center for Supercomputing R&D
- hblim@csrd.uiuc.edu Univ. of Illinois at Urbana-Champaign
-
-