home *** CD-ROM | disk | FTP | other *** search
- Path: informatik.tu-muenchen.de!fischerj
- From: fischerj@informatik.tu-muenchen.de (Juergen "Rally" Fischer)
- Newsgroups: comp.sys.amiga.programmer
- Subject: Re: C2P
- Date: 17 Jan 1996 18:16:53 GMT
- Organization: Technische Universitaet Muenchen, Germany
- Distribution: world
- Message-ID: <4djeel$365@sunsystem5.informatik.tu-muenchen.de>
- References: <38231876@kone.fipnet.fi> <4cuhbv$l14@oreig.uji.es> <38231974@kone.fipnet.fi> <4de4s1$kjj@oreig.uji.es>
- NNTP-Posting-Host: hphalle5.informatik.tu-muenchen.de
- Originator: fischerj@hphalle5.informatik.tu-muenchen.de
-
-
- In article <4de4s1$kjj@oreig.uji.es>, ii202@rossegat.uji.es (Jorge Acereda Macia) writes:
- |> Organization: Universitat Jaume I. Castell≤ de la Plana. Spain
- |> Lines: 27
- |> Distribution: world
- |> Message-ID: <4de4s1$kjj@oreig.uji.es>
- |> References: <38231876@kone.fipnet.fi> <4cuhbv$l14@oreig.uji.es> <38231974@kone.fipnet.fi>
- |> NNTP-Posting-Host: @rossegat.uji.es
- |> X-Newsreader: TIN [version 1.2 PL2]
- |>
- |> Jyrki Saarinen (jsaarinen@kone.fipnet.fi) wrote:
- |>
- |> > > 20 ms? Using scrambled chunky buffer I get 320x200x8 in 25.7
- |> > > on 030@50 (slow :-( ). Well, I keep multitasking enabled using a 127
- |> > > TaskPri, so interrupts should suck the cache... But 20 ms. for
- |> > > 320x256... Could you give some hints?
- |>
- |> > I was just speculating. Copying a 320x256x8 data area from
- |> > fast to chip ram takes about 20ms. Now, if three c2p passes
- |> > could be done "free" on 030/50 and the last pass with
- |> > the blitter, it would be great.
- |>
- |> Using a scrambled buffer, I assume...
- |>
- |> > I dont think your systemfriendliness slows down..
- |> > Do you use the blitter?
- |>
- |> No blitter. 320x200 with CPU only, takes 25.7 ms, beeing the first two
- |> passes 5.something ms and the other two+writes to chip ram the other 20 ms.
- |> (030@50 times) It's pretty slow. I think three totally free passes are not
- |> possible on 030, but the times should not be more than 22 or 23 ms (?).
- |>
- |> So, how much would require the last pass with the blitter? Will we
- a blitter pass on 320x256x8planes needs about 0.04 sec.
-
- |> see full screen 25 fps doom on 030?
-
- well, if the copying to vram (assuming free 2pss action) needs a frame
- there's not much left.
-
- 25fps are rather ratings for a well cached 486-66 (no rather 486-80).
-
- but maybe a 030-50 with good mem interface doing 320x160 can go 3 frames,
- about 1/2 frame copy, 1.5 frames for raw-mapping (assuming 20cyle inner loop.
- shading ?), so 1 frame left for polygon 1st-outer-loops+2nd-outer,
- rotation, z....
-
- well, a A1200+ with 030-40 and 4 frame doom, aaaah :)
-
- |>
- |> Greets,
- |> Jorge Acereda (ii202@rossegat.uji.es)
- ------------------------------------------------------------------------
- fischerj@Informatik.TU-Muenchen.DE (Juergen "Rally" Fischer) =:)
-
-