home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!zaphod.mps.ohio-state.edu!cis.ohio-state.edu!ucbvax!cgl!cgl.ucsf.edu!hatton
- From: hatton@socrates.ucsf.edu (Tom Hatton)
- Newsgroups: comp.os.os2.advocacy
- Subject: Re: SGI-GL in Windows NT??
- Message-ID: <hatton.726700891@cgl.ucsf.edu>
- Date: 10 Jan 93 21:21:31 GMT
- References: <dlcogswe.726380388@vela> <8368@lib.tmc.edu> <1993Jan8.174312.26479@solaris.rz.tu-clausthal.de> <hatton.726530814@cgl.ucsf.edu> <930109083021@rgam.sc.ti.com> <hatton.726615957@cgl.ucsf.edu> <dlcogswe.726642704@vela>
- Sender: news@cgl.ucsf.edu (USENET News System)
- Organization: UCSF Computer Graphics Lab
- Lines: 47
-
- dlcogswe@vela.acs.oakland.edu (Dan Cogswell) writes:
-
- >hatton@socrates.ucsf.edu (Tom Hatton) writes:
-
- >>Thanks for providing a ballpark for comparisons, but apparently Mr. Cogswell
- >>has all the time in the world. For myself, having brought an SGI machine
- >>to a crawl (4GX? Don't remember the model) doing molecular modeling, I
- >>suspect that I would not be happy with the performance of NT with SGI -GL
- >>in software on a 486.
-
- >Molecular modelling? How many polygons and what shading algorithm??
- >Seems to me you brought the CPU on the machine to a crawl, not the
- >graphics engine. Correct me if I'm wrong. If that's the case, you
- >wouldn't see much difference between to dissimilar machines with
- >comparable CPUs.
-
- >My current project (radiosity) spends most of it's time in computation,
- >even with the un-optimized gouraud shade routine I've written. Could
- >you explain to me how having polygons in hardware would speed things up
- >significantly?
-
- >Of course, I'm ignorant to what else the Irises do in hardware, so
- >please fill me in.
-
- I know neither the number of polygons, nor the shading algorithm, since
- I was merely using the software, not writing it; and admittedly there
- was likely an element of serious CPU computation as a factor. My point is,
- that the same thing (animation of a 3D protein model) on 486, or even
- several 486's, without the extra video goodies the Iris (again I would
- have to check details for the 4GX, etc.) provides in hardware, it would
- be ludicrous.
-
- If serious hardware comes along that helps this out, then fine; but your
- original comment seemed to be "NT has the SGI graphics lib in software,
- and that's enough" and I take exeception to that, unless sufficient
- hardware is there also. Given the time frame for getting the *next*
- version of NT out (ie., the one after NT 3.1) and the lag time for getting
- people to use NT, the SGI stuff, etc., I think the alternatives will still
- look more attractive. But if this feature, as distant as it now seems, is
- enough to make you go for NT, then fine. If it works for you, fine. But
- allow me to remain sceptical.
-
-
- --
- Tom Hatton
- hatton@cgl.ucsf.edu
- (415)-476-8693
-