Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!uwm.edu!linac!pacific.mps.ohio-state.edu!zaphod.mps.ohio-state.edu!samsung!uunet!cbmvax!daveh From: daveh@cbmvax.commodore.com (Dave Haynie) Newsgroups: comp.sys.amiga.misc Subject: Re: Few questions/observations about the A3000 Keywords: KS 2.0 Message-ID: <21193@cbmvax.commodore.com> Date: 3 May 91 03:24:36 GMT References: <1991Apr24.214121.4631@javelin.sim.es.com> <1991Apr25.154002.4471@daffy.cs.wisc.edu> <1991Apr26.202103.22105@javelin.sim.es.com> <1991Apr30.160504.22656@daffy.cs.wisc.edu> Reply-To: daveh@cbmvax.commodore.com (Dave Haynie) Organization: Commodore, West Chester, PA Lines: 64 In article <1991Apr30.160504.22656@daffy.cs.wisc.edu> pochron@rt1.cs.wisc.edu (David Pochron) writes: >In article <1991Apr26.202103.22105@javelin.sim.es.com> blgardne@javelin.sim.es.com writes: >>How are you getting SetCPU to load KS into 16 bit RAM? >Simple...Boot the KS from a disk, either from a "kickfile" or a KickStart >disk itself. It loads it into C00000 memory (which, I know, is slightly >slower than "real" fast 16 bit RAM, but I am taking that into account) and >as long as you never do a FASTROM when you reboot, it stays there. Sure does. Because I called it FASTROM, I felt obligated to have SetCPU make it's best attempt at finding some 32 bit RAM and using it. But the ROM translation stuff should, in theory, work out of any memory. Chip RAM or the Chip bus $00C00000 RAM is used for KICKROM boots, simply because, after RESET, all RAM goes away, and that's the easiest to get back -- no special autoconfiguration magic to try. While SetCPU looks specifically for A26x0 32 bit RAM, on all other systems it'll use the first pool of Fast RAM it can find for its FASTROM. >>You're mistaken here. There are two distinct things giving you the >>speed up. First is the 32 bit wide memory. Second is the 80 nS RAM vs >>the 250 nS ROM. When KS is in RAM you don't have the wait states the ROM >Hmmm, That is interesting. I thought ROMs were just like static RAM - no >wait states. Well, if that is the case, then you have enlightened me. Static RAM and ROM do follow similar principles. Both are indeed statically addressed. Neither are every "no wait states" on an arbitrarily fast machine. In both cases, there's always a speed versus size tradeoff. You can get tiny ROMs, generally used for control or microROM, in the 1Kbit range, that go 10ns or so. When you talking about 4Mbits, you're lucky to get 150ns or thereabouts. Though they do get faster every year. >>I understood the marketing reasons as to why kickstart had to go to ROM >>for the A500/A2000. In _my_ opinion, one of the things I really missed >>was the flexibility of the kickstart design. It is too bad the market- >>place is too stupid to appreciate it. >Sorry, I can't buy that. I have a 1000 downstairs and it is a pain in the >butt to have to load KickStart all the time. For your primary boot, KickStart off a floppy is real annoying. We'd be laughed out of the market if a system like the A3000 had to do it that way. From hard disk, you barely notice it. I run 2.0 similarly on an A2500, via a little Startup-Sequence nonsense, and it's nearly as fast booting as the A3000. >I don't see what the big deal about KS on disk is...KS upgrades (for the >general public...I don't have 2.0, do you?) come no more often as they >come on ROM chip. I don't see it as a big deal either. Then again, I used to build (stuff and solder up) A2000 motherboards by hand (our technicians have since learned about the wave solder machine downstairs, so that's not longer necessary). But some people just object to changing ROMs. And new ROMs will always be a tad more expensive than one more floppy. So there's something to be said about making 2.0 available on disk. Whether they will or not is anyone's guess. > David M. Pochron | Transparent DWV pipe: For the man who wants to -- Dave Haynie Commodore-Amiga (Amiga 3000) "The Crew That Never Rests" {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy "That's me in the corner, that's me in the spotlight" -R.E.M.