On Tue, Jan 27, 2009 at 5:36 AM, mabshoff <mabsh...@googlemail.com> wrote: > > > > On Jan 26, 12:48 pm, William Stein <wst...@gmail.com> wrote: >> Hi, > > <SNIP> > >> > Other issues are performance; >> > running a Sage server in a VMware session is noticeable slower than >> > connecting to an on-line Linux version. >> >> For *raw computations* (cpu bound code, e.g., computing determinants, >> solving systems, etc.) VMware Sage is likely mostly going >> to be faster than a native windows port, especially on modern >> processors. > > I *highly* doubt that when "native port == MSVC". MSVC handily beats > gcc on spec and is nearly as fast as icc. I know, it is benchmarks, > damn benchmarks and lies, but my personal experience matches that. > MS's compiler group has a lot of high carat people and while > performance of VC6 was terrible loads of things have changed in the > last decade :). VS 2005 and higher have loads of tools that are well > integrated and IMHO not found anywhere else in such a high degree of > integration and consistency. >
In theory native MSVC should do as well or better, but in practice this still depends a lot on people actually tuning actual code. I'm sure I can write code that works well in Linux and not so well on Windows (e.g., use fork and pseudotty's everywhere... :-) ). That said, having eMPIRe at the core, which has very good MSVC support by Brian Gladman, will be very good for Sage on Windows. -- William --~--~---------~--~----~------------~-------~--~----~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to sage-devel-unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://www.sagemath.org -~----------~----~----~----~------~----~------~--~---
