Article View: alt.alien.research
Article #1012Re: SETI@home doesn't finish Data Units
From: Dave Storey
Date: Thu, 24 Jun 1999 00:00
Date: Thu, 24 Jun 1999 00:00
56 lines
3711 bytes
3711 bytes
In article <lCdc3.2431$Lk4.28048@news4.mia>, Jim K <star@watcher.invalid> writes >Dave, >Would the type of video card, you have, make any difference in the time >spent processing the graphics? Marginal .. the bitblt to the graphics memory is not where all the time goes, its deciding which stupid pixels are on/off in what colours, and whether they are hidden by the rest of the 3D bar chart. That sort of graphics is CPU intensive. Note that the video card doesn't even get a look in when the Application is minimised (however the 'colour depth' setting, eg 256 colours, 16 bit, or 24/32 bit DOES have an effect, since that affects the size of the graphics block that the App has to build, what u-soft is pleased to call 'the device context'). >What have you heard about the vcache settings? Nothing .. should I? >My system is a P233 MMX IE5, only wish I had a P2 or P3. With no activity, >other than the screensaver blanked, the best I could hope for is around >24hrs/block. Sounds reasonable, +/- an hour or two. My numbers assume enough memory (32Meg plus), and the Pentium is not quite as swift at FP maths if I remember right. You have to remember that '233' is more of a wish than a promise .. some 233 systems run at about 210, and some at 250. SETI have got back some bad results from, probably, over-clocked machines .. you can wind a 233 up to rather more, and often get away with it, UNTIL you run the FPU, or cache, or <part X>, flat out for 24 hours, when it overheats and starts missing the odd bit here and there. GIMPS/Primenet, for instance, runs a 1hr stress test on your machine before it lets you loose to look for primes (I have NO idea what it does if your machine fails) - S@H seems to be relying on duplication - pretty safe bet at the moment! >Jim > >Dave Storey <dave@quik.demon.co.uk> wrote in message >news:syBGnDAanVc3Ewy4@quik.demon.co.uk... >> >> Actually, having measured it, it's more like 2/3rds to 3/4s 8<. >> >> On an 'idle' system, with no graphics, and a Gen-U-Ine Intel Pentium >> Pro, 2 or 3, you should manage about 5k/clockspeed (in Mhz), so a 233 >> should do a block in ~21 hours. If it's taking 35, then your system is a >> bit busy 8>. >> >> Rgds >> Dave Storey > > Rgds Dave Storey
Technical Details
Message-ID:
<1hewTAAn+Wc3Ewgl@quik.demon.co.uk>Path:
arc1-ca.pugleaf.net!unknown!not-for-mailReferences:
<7i2hnp$hbc$1@front7.grolier.fr> <7i49ij$g7i$1@lure.pipex.net> <374d69d0.36972401@news.earthlink.net> <MPG.11b1c56b53cc360698ad87@news.demon.co.uk> <7i8r76$fdb$1@news4.svr.pol.co.uk> <3748d6c9.88890764@news.earthlink.net> <37617CEB.251DD346@cwo.com> <7k9f8b$5et$1@MANNET-3800-2.mcb.net> <7k9psh$pe8$1@helix.cs.uoregon.edu> <37713a2f.70843597@news.magmacom.com> <m%2c3.1538$Gg3.17188@news3.mia> <3771fb17.120218034@news.magmacom.com> <pwcc3.2217$mj5.23454@news1.mia> <syBGnDAanVc3Ewy4@quik.demon.co.uk> <lCdc3.2431$Lk4.28048@news4.mia>Show Full Headers
Complete Article Headers
X-Google-Language: ENGLISH,ASCII-7-bit X-Google-Thread: 10857a,fe0b62700be4323b X-Google-Attributes: gid10857a,public X-Google-Thread: fdc67,fe0b62700be4323b X-Google-Attributes: gidfdc67,public X-Google-Thread: f98ed,fe0b62700be4323b X-Google-Attributes: gidf98ed,public X-Google-Thread: 113975,fe0b62700be4323b X-Google-Attributes: gid113975,public From: Dave Storey <dave@quik.demon.co.uk> Subject: Re: SETI@home doesn't finish Data Units Date: 1999/06/24 Message-ID: <1hewTAAn+Wc3Ewgl@quik.demon.co.uk> X-Deja-AN: 493146143 X-NNTP-Posting-Host: quik.demon.co.uk:194.222.100.178 References: <7i2hnp$hbc$1@front7.grolier.fr> <7i49ij$g7i$1@lure.pipex.net> <374d69d0.36972401@news.earthlink.net> <MPG.11b1c56b53cc360698ad87@news.demon.co.uk> <7i8r76$fdb$1@news4.svr.pol.co.uk> <3748d6c9.88890764@news.earthlink.net> <37617CEB.251DD346@cwo.com> <7k9f8b$5et$1@MANNET-3800-2.mcb.net> <7k9psh$pe8$1@helix.cs.uoregon.edu> <37713a2f.70843597@news.magmacom.com> <m%2c3.1538$Gg3.17188@news3.mia> <3771fb17.120218034@news.magmacom.com> <pwcc3.2217$mj5.23454@news1.mia> <syBGnDAanVc3Ewy4@quik.demon.co.uk> <lCdc3.2431$Lk4.28048@news4.mia> X-Complaints-To: abuse@demon.net X-Trace: news.demon.co.uk 930182119 nnrp-01:25961 NO-IDENT quik.demon.co.uk:194.222.100.178 Organization: Heap, rather than Stack. MIME-Version: 1.0 Newsgroups: alt.alien.research,alt.astronomy,fr.sci.astronomie,sci.astro.amateur