Newsgroups: comp.os.msdos.djgpp,comp.os.msdos.programmer,alt.msdos.programmer,alt.games.programming From: "William A. Barath" Subject: Re: High resolution timer problems.. In-Reply-To: <65m5mt$ka9$1@news2.xs4all.nl> Content-Type: TEXT/PLAIN; charset=US-ASCII Message-ID: Sender: news AT freenet DOT victoria DOT bc DOT ca (News Manager) Organization: Victoria Freenet Association References: <65m5mt$ka9$1 AT news2 DOT xs4all DOT nl> Mime-Version: 1.0 Date: Fri, 28 Nov 1997 12:08:42 GMT Lines: 30 To: djgpp AT delorie DOT com DJ-Gateway: from newsgroup comp.os.msdos.djgpp Precedence: bulk On 28 Nov 1997 rkramer AT xs4all DOT nl wrote: |In a discussion with a friend of mine, he was trying to convince me that |it's possible to blit a full 800x600 screen in less than a millisecond. I |tried to tell him otherwise, but couldn't quite get my point through :) | |Can anyone tell me how I could calculate the time it would take to blit a |full VESA 0x103 screen, using a 486DX2 66 MHz? Is this an easy task, or a |very difficult one? Do I have to keep in mind: Well lesee here. dx2/66 means a best bus speed of 33Mhz, with 32-bit bandwidth, yielding a *possible* 132Mbyte/sec throughput, assuming zero waitstates and zero hazards. Now an 800x600 screen on a local bus graphics adapter just might be zero waitstate, so lets continue giving your friend the benefit of the doubt. Lets also assume a 256-color screen because otherwise we are guaranteed large bus waits for VGA memory. So we are going to blit 800*600 bytes, approx .5 meg, with 132Mbytes/sec of bandwidth. Assuming that load and store will take the same amount of bandwidth, we can blit 66 megs through his bus per second. Now to do it in a millisecond, we need 1000 times .5meg times two, or a gig/second of bandwidth. But lets again be charitable and assume that he has hardware that nobody else on the planet has, and give the point to him. Not. -Wil Barath "I feel as though I see my pen to write" aka WseM homepage: http://victoria.tc.ca/~wi534