home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!cbmvax!daveh
- From: daveh@cbmvax.commodore.com (Dave Haynie)
- Newsgroups: comp.sys.amiga.hardware
- Subject: Re: ECS, AGA bandwidth?
- Message-ID: <37810@cbmvax.commodore.com>
- Date: 11 Dec 92 16:46:25 GMT
- References: <MJL.92Nov27055025@dino.ph.utexas.edu>
- Reply-To: daveh@cbmvax.commodore.com (Dave Haynie)
- Distribution: comp
- Organization: Commodore, West Chester, PA
- Lines: 30
-
- In article <MJL.92Nov27055025@dino.ph.utexas.edu> mjl@dino.ph.utexas.edu (Maurice J. LeBrun) writes:
-
- >A friend of mine recently bought a 486/localbus machine and says he can get
- >100 Mbits/sec between the video memory and CPU. Sounds to me that this is a
- >bit faster than ECS but I'm not sure -- does anyone have the figures for ECS
- >bandwidth for an A3000?
-
- 100 Mbits/sec == 12.5 MB/s. The A3000's CPU bandwidth to Chip RAM is up to
- 7 MB/s, depending on how much time is being stolen by the ECS chips (could be
- none, could be lots).
-
- >Also, I've seen about a dozen times that AGA is 4X faster than ECS for the
- >highend machines. I believe this is due to doubling the data path (32 bits
- >instead of 16) and doubling the internal clock speed (4x7.16 MHz instead of
- >2x7.16 MHz). Is this correct?
-
- Kind of. The bus is now 32-bits wide. The Lisa chips does run from the 28MHz
- clock directly, but that's primarily used for nice, clean pixel transitions.
- The extra bandwidth is in video fetches. In normal operation, Lisa grabs
- two 32-bit words in the time Denise would grab one 16-bit word. This works
- a bit like the A3000's burst mode.
-
-
-
- --
- Dave Haynie / Commodore Technology, High-End Amiga Systems Design (cool stuff)
- "The Crew That Never Rests" {uunet|pyramid|rutgers}!cbmvax!daveh BIX: hazy
- SCIENCE: "I'll believe it when I see it"
- RELIGION: "I'll see it when I believe it"
-
-