ralphw@IUS3.IUS.CS.CMU.EDU (Ralph Hyre) (03/09/88)
In article <4744@ecsvax.UUCP> urjlew@ecsvax.UUCP (Rostyk Lewyckyj) writes: >In article <3434@cbmvax.UUCP>, daveh@cbmvax.UUCP (Dave Haynie) writes: >> in article <390@ontenv.UUCP>, norm@ontenv.UUCP (Norman S. Soley) says: >> > Keywords: 68030 >> > Xref: cbmvax comp.sys.m68k:830 comp.sys.misc:1315 >> >> > The subject line says it all. But I'll repeat it. Is anyone aware of a >> > 68030 based UNIX system which is available now or in the near future? >> >so close upon us, I can't resist ;-). I think it's time for us to >start a thread on comp.sys.m68k and comp.sys.amiga on designing a new Amiga [Ranger 3500, C3000] or some other catchy code name (C2001, C2010)?. ... >Actually for a while there was a serious discussion in, of all places, >The Computer Shopper, on a public domain design for an open MAC. >This was at a time when Jobs at Apple was insisting on a closed box >with 256-512K of memory. I think that this input did influence the >design of the MAC. They chose a standardized pd buss, SCSI disk OK, the Amiga 3000 should be a 16Mhz 68030/882 machine. It'll be interesting to see what the graphics hardware will look like in the next generation of Amiga. IBM PC-style slots would be a good choice, although a provison for Microchannel or NuBUS boards would be a good idea, since they're the up-and-coming busses. There should be a provision for running Mac software and reading 800k and 1.6M Mac disks [One can imagine Mac Toolbox compatible machines (use the same [documented] dispatch table entry points & parameters, although the code must be different) similar to PC BIOS clone machines]. If there's existing software out there that will work, then it establishes a compatibility standard which software can be rated on: ('Foobaz' runs on Mac+, SE, Mac ][, and and the Amiga with the MacAmiga software, but MacST, for example) (A third party product is OK, but it's better if Commodore-Amiga views it as a strategic product and supports it in some way.) -- - Ralph W. Hyre, Jr. Internet: ralphw@ius2.cs.cmu.edu Phone:(412)268-{2847,3275} CMU-{BUGS,DARK} Amateur Packet Radio: N3FGW@W2XO, or c/o W3VC, CMU Radio Club, Pittsburgh, PA
cmcmanis%pepper@Sun.COM (Chuck McManis) (03/10/88)
In article <1069@PT.CS.CMU.EDU> ralphw@IUS3.IUS.CS.CMU.EDU (Ralph Hyre) writes: > OK, the Amiga 3000 should be a 16Mhz 68030/882 machine. It'll be > interesting to see what the graphics hardware will look like in the > next generation of Amiga. Well these are the new custom chips I would like to see ... Mary - Mary is a serial interface chip optimized for MIDI applications. She has two 31.25K baud serial ports and a 16 bit timer that provides .1 millisecond resolution. She has a 32 bit bus and when you read a word you get the two MIDI data bytes and a 16 timestamp. She will optionally generate 'double zero' bytes when the timer rolls over. (About every 6.5 seconds) Paula - She has had here counters reworked so that she can generate 16 bit samples (8 channels) at 48 Khz. Additionally, the divisor channels are replaced with additive synthesis dividers which give linear control over the frequency. Agnus - Of course agnus addresses the full 2Meg of CHIP ram. Victoria - Victoria is a new chip with 20 floating point registers organized as a 4 X 4 array. Built in trancendental support allows here to do a vector transformation in 21 cycles. Victoria has two DMA channels, a source and destination. When started she will burst fetch "n" elements (1 - 4) from the source DMA pointer, apply the transformation and store "m" elements (1 - 4) in the destination DMA address. Further there is a number of "points" register that limits the total number of vectors fetched. Denise - Well Denise is of course expanded to 8 bits/pixel and 16M colors. Display modes are bumped up to 800 X 600 in color, 1600 X 1200 in monochrome. The C1180 is Commodores 19" color RGB multisync monitor. >There should be a provision for running Mac software and reading 800k >and 1.6M Mac disks [One can imagine Mac Toolbox compatible machines >(use the same [documented] dispatch table entry points & parameters, >although the code must be different) similar to PC BIOS clone machines]. >If there's existing software out there that will work, then it establishes a >compatibility standard which software can be rated on: Fooey, if I wanted a stupid Mac I would have bought one. The only reason I can see you wanting this is to let you run software that the Mac currently runs. I envision that the Mac and PC people will port their stuff to the Amiga because it offers the most bang for the buck and the nicest environment. That and the 10,000,000 Amigas in the world that C/A will have sold. +--------------------------------------------------------------+ | System ALERT : Drug use detected! | | Don't believe for a *second* any of this is in the works! | | Press Left Button to Continue Price Right Button to Reboot | +--------------------------------------------------------------+ --Chuck McManis uucp: {anywhere}!sun!cmcmanis BIX: cmcmanis ARPAnet: cmcmanis@sun.com These opinions are my own and no one elses, but you knew that didn't you.
Chad_The-Walrus_Netzer@cup.portal.com (03/10/88)
In article <1069@PT.CS.CMU.EDU> (Ralph Hyre) writes: >OK, the Amiga 3000 should be a 16Mhz 68030/882 machine. It'll be >interesting to see what the graphics hardware will look like in the >next generation of Amiga. Why just a 16 Mhz? Commodore should go right for the gusto and produce the highest clockspeed Amiga possible (I have no idea what this would be). This would gain them some attention. >IBM PC-style slots would be a good choice, although a provison for >Microchannel or NuBUS boards would be a good idea, since they're the >up-and-coming busses. NO! NO! NO! NOT the NuBUS!!!!! It is pure crap for a an highspeed/highpower computer, or even a lower speed/lesser power one! At least, Apple's implementation of it is... A good BUS is ESSENTIAL to the throughput of a machine, and the NuBUS is a MEGA blunder in this respect... It will only slow things down. I recently heard an interesting comment from an author (perhaps THE author) of JFORTH. I don't remember his name, but he said that as computers are getting more advanced, we need to dump the idea of having this IBM type chassis, where all the board are allinged like books on a book shelf. He stated that we should go for a more CRAY-like architecture, which will reduce interference, and allow for more speed, power, etc. It sounded like a GREAT idea to me, but since it won't like as nice, I doubt it will happen for a while... Sorry, if I slightly misquoted the guy, but I think I've expressed the gist of what he said. If not, it's my fault... Anyway, that's how I see it... Chad 'The_Walrus' Netzer ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
eric@hector.UUCP (Eric Lavitsky) (03/11/88)
In article <1069@PT.CS.CMU.EDU> ralphw@IUS3.IUS.CS.CMU.EDU (Ralph Hyre) writes: >In article <lots of places> various people write... (much deleted) > >OK, the Amiga 3000 should be a 16Mhz 68030/882 machine. It'll be >interesting to see what the graphics hardware will look like in the >next generation of Amiga. Fine, I'll buy it... the next generation of graphics hardware will most likely be an extension of the current architecture: higher resolution, more color, greater addressing range, ability to blit more data at a time at higher speeds etc. They should remain as compatible with the existing machines as possible. >IBM PC-style slots would be a good choice, although a provison for >Microchannel or NuBUS boards would be a good idea, since they're the >up-and-coming busses. No, no no! - Don't ever even think this; it would be a huge mistake! What the next Amiga really needs is a supported 32 bit extension to the current Zorro II bus architecture! > - Ralph W. Hyre, Jr. Eric ARPA: eric@topaz.rutgers.edu "Lithium is no longer available UUCP: ...{wherever!}ulysses!eric on credit..." ...{wherever!}rutgers!topaz!eric - from Buckaroo Banzai SNAIL: 34 Maplehurst Ln, Piscataway, NJ 08854
lphillips@lpami.van-bc.UUCP (Larry Phillips) (03/12/88)
In article <912@cfa.cfa.harvard.EDU> ward@cfa.harvard.EDU (Steve Ward) writes: >I heard a rumor to the effect that there exists a Public Domain M680x0 >assembler and/or C Compiler available in source form. Charlie Gibbs of vancouver BC has written an assembler and the source for it is available. It is on Fred Fish #110 along with PDC. Charlie is now up to version 1.06, which will be posted to Compuserve within a short time. In addition, Jeff Lydiatt has been doing a lot of work on PDC (now renamed to Publicly Redistributable C, or PDC :-) ). He says he is getting close to having it ready for distribution. The one on FF #110 does suffer some severe bugs. -larry -- The transistor is a curiosity, and will never amount to much. -- Mr. Stringer, Basic Electronic Instructor, RCAF, 1962. +--------------------------------------------------------------------+ | // Larry Phillips UUCP: lphillips@lpami.van-bc.UUCP | | \X/ or: {ihnp4!alberta!ubc-vision,uunet}!van-bc!lpami!lphillips | | COMPUSERVE: 76703,4322 | +--------------------------------------------------------------------+
sean@ms.uky.edu (Sean Casey) (03/18/88)
In article <1069@PT.CS.CMU.EDU> ralphw@IUS3.IUS.CS.CMU.EDU (Ralph Hyre) writes: >OK, the Amiga 3000 should be a 16Mhz 68030/882 machine. It'll be >interesting to see what the graphics hardware will look like in the >next generation of Amiga. Sixteen Megahertz????? That's like getting a Porsche 911 and putting a Volkswagen engine in it. Surely the '030 will run at a faster clock speed than 16 MHz? The '020 runs at 25. And the better damn well get rid of flicker when running at interlace resolution. Sean -- *** Sean Casey sean@ms.uky.edu, sean@ukma.bitnet *** The Empire Maniac {rutgers,uunet,cbosgd}!ukma!sean *** University of Kentucky / Lexington Kentucky / USA *** "Ludo... FRIEND!"
david@ms.uky.edu (David Herron -- Resident E-mail Hack) (03/21/88)
In article <912@cfa.cfa.harvard.EDU> ward@cfa.harvard.EDU (Steve Ward) writes: >>I heard a rumor to the effect that there exists a Public Domain M680x0 >>assembler and/or C Compiler available in source form. While it's not PD, it's free and source comes along with it along with a strange liscence requiring any re-distribution to include the source. I'm of course talking about the GNU C Compiler and GNU Assembler. Both are somewhat portable and such. They work on 68000 class machines however. I haven't looked into what would need to be fixed up to use 'em on an Amiga. Of course the assembler would need to be taught about the amiga linker format. Or possibly you could use somebody else's assembler? (In which case the C Compiler would have to be taught about the assembler format :-)). We've got both here in ~ftp/archive/GNU, ftp to a.ms.uky.edu. -- <---- David Herron -- The E-Mail guy <david@ms.uky.edu> <---- or: {rutgers,uunet,cbosgd}!ukma!david, david@UKMA.BITNET <---- <---- "Oh, I dunno -- I think Sean would be rather tasty!" -- Becky
keithd@cadovax.UUCP (Keith Doyle) (03/22/88)
In article <8608@g.ms.uky.edu> sean@ms.uky.edu (Sean Casey) writes: >And the better damn well get rid of flicker when running at interlace >resolution. But not at the expense of NTSC compatibility!!!! (in other words, it better do both, or I'd rather stick with interlace). All I really want is more CHIP ram. Keith Doyle # {ucbvax,decvax}!trwrb!cadovax!keithd Contel Business Systems 213-323-8170
scott@applix.UUCP (Scott Evernden) (03/23/88)
In article <2003@cadovax.UUCP> keithd@cadovax.UUCP (Keith Doyle) writes: > > All I really want is more CHIP ram. Like maybe the entire address space? (i.e., Why would/must there even be such a thing as CHIP ram??) -scott
flaig@cit-vlsi.Caltech.Edu (Charles M. Flaig) (03/23/88)
In article <682@applix.UUCP> scott@applix.UUCP (Scott Evernden) writes: >In article <2003@cadovax.UUCP> keithd@cadovax.UUCP (Keith Doyle) writes: >> >> All I really want is more CHIP ram. > >Like maybe the entire address space? > >(i.e., Why would/must there even be such a thing as CHIP ram??) > >-scott Having separate Chip RAM allows the co-processors to run at full speed using Chip RAM, and the 680x0 to run at full speed using "Fast" RAM. The so-called "Fast" RAM isn't actually any faster, but programs executing out of it run faster since they don't have to share access cycles with the co-processors. This is especially important when your screen is using overscan and/or a large number of bitplanes. So, it's not *essential* to have separate Chip RAM (an unexpanded A1000 has only Chip RAM for instance), but for my money it's worthwhile! ______________________________________________________________________________ ___ , , ,;,;;;, / Y /| /| Charles Flaig ;/@-@\; | |/ __, ,__ |/ flaig@csvax.caltech.edu | ^ | | /^\ / | | | / /\ /\ \=/ \____/| \_/|_/\_/ \_/ \_\/_/_/_/ "What, you think they PAY me for this?"
lel@wuphys.UUCP (Lyle E. Levine) (03/24/88)
In article <682@applix.UUCP> scott@applix.UUCP (Scott Evernden) writes: >In article <2003@cadovax.UUCP> keithd@cadovax.UUCP (Keith Doyle) writes: >> >> All I really want is more CHIP ram. > >Like maybe the entire address space? > >(i.e., Why would/must there even be such a thing as CHIP ram??) > >-scott Personally, I LIKE having separate CHIP & FAST RAM. Putting stuff in FAST RAM eliminates bus contention. I only put what I have to in CHIP. However, I would like more CHIP RAM. An idea! Hmm..., off hand, I can't think of an easy way to do this, but could this be made configurable? Say, I'd like 2 MEG chip (assuming Aggy can handle this!). Or .5 Meg chip. Just let the system configure itself to this. Any ideas on a method for this? ========== IBM is a Division of Sirius Cybernetics Corporation "their fundamental design flaws are completely hidden by their superficial design flaws." - "So Long And Thanks For All The Fish" Lyle Levine: Paths -> ihnp4!wuphys!lel uunet!wucs!wuphys!lel
charles@hpcvca.HP.COM (Charles Brown) (03/24/88)
>>And the better damn well get rid of flicker when running at interlace >>resolution. >> Sean Casey >But not at the expense of NTSC compatibility!!!! (in other words, it >better do both, or I'd rather stick with interlace). All I really >want is more CHIP ram. > Keith Doyle It is probably not possible to satisfy everyone. I would rather have high resolution with no interlace. I am willing to give up NTSC compatibility for most computer use. The only way to satisfy both your wants and mine is to provide NTSC mode and Hi-Resolution mode. Of course, you could not have both at once. charles@hp-pcd
gore@eecs.nwu.edu (Jacob Gore) (03/25/88)
/ comp.sys.amiga / lel@wuphys.UUCP (Lyle E. Levine) / Mar 23, 1988 / >Personally, I LIKE having separate CHIP & FAST RAM. Putting stuff >in FAST RAM eliminates bus contention. So, all we need is dual-ported memory. And a recipie by which Commodore could use it and yet keep the machine cost under $5K... Jacob Gore Gore@EECS.NWU.Edu Northwestern Univ., EECS Dept. {oddjob,gargoyle,ihnp4}!nucsrl!gore
daveh@cbmvax.UUCP (Dave Haynie) (03/25/88)
in article <8608@g.ms.uky.edu>, sean@ms.uky.edu (Sean Casey) says: > That's like getting a Porsche 911 and putting a Volkswagen engine in it. > Surely the '030 will run at a faster clock speed than 16 MHz? The '020 > runs at 25. The current '030s are in 16.67Mhz and 20Mhz versions. I think Moto has announced 25Mhz planned before the end of the year, and I for one would certainly like to see a 30Mhz version some day. On the other hand, for the same speed clock, the 68030 can run memory cycles over twice the speed of the 68020/68851 combination.... > *** Sean Casey sean@ms.uky.edu, sean@ukma.bitnet -- Dave Haynie "The B2000 Guy" Commodore-Amiga "The Crew That Never Rests" {ihnp4|uunet|rutgers}!cbmvax!daveh PLINK: D-DAVE H BIX: hazy "I can't relax, 'cause I'm a Boinger!"
cmcmanis%pepper@Sun.COM (Chuck McManis) (03/25/88)
Keith Doyle wrote : >> But not at the expense of NTSC compatibility!!!! Then Charles Brown wrote : >>I would rather have high resolution with no interlace. >>Of course, you could not have both at once. With the FlickerFixer in a 2000 you get both at once. The FlickerFixer has a multisync monitor output on it and the Amiga output is still available so you can dedicate that output to the genlock'd video you are sending to the VCR. --Chuck McManis uucp: {anywhere}!sun!cmcmanis BIX: cmcmanis ARPAnet: cmcmanis@sun.com These opinions are my own and no one elses, but you knew that didn't you.
avery@puff.cs.wisc.edu (Aaron Avery) (03/25/88)
In article <3512@cbmvax.UUCP>, daveh@cbmvax.UUCP (Dave Haynie) writes: > The current '030s are in 16.67Mhz and 20Mhz versions. I think Moto has > announced 25Mhz planned before the end of the year, and I for one would > certainly like to see a 30Mhz version some day. You got this right. However, Motorola themselves have 68030's running at 30Mhz (and 68882's!) NOW. I've seen one. They're making a VMEmodule which is a 'Monoboard Microcomputer' with 25 and 30Mhz clocks. Probably one of the best size/performance ratios to date. > On the other hand, for the same speed clock, the 68030 can run memory cycles > over twice the speed of the 68020/68851 combination.... Over twice? This must be only in cases of 0ns internal MMU translation. The '030 can have a 2-clock memory cycle, while the '020 is limited to a 3-clock memory cycle, so there's this advantage up front. The other big advantage is the added internal data cache in the '030. Then there's the internal pipelining, to allow instruction fetch, address translation, effective address calculation, and arithmetic operation all functioning simultaneously (hope I didn't stretch it one too far, but you get the point). Basically, this is one FINE processor, and I hope to be playing with one in the near future. Aaron Avery (avery@puff.cs.wisc.edu) ({seismo,caip,allegra,harvard,ihnp4}!uwvax!puff!avery)
daveh@cbmvax.UUCP (Dave Haynie) (03/29/88)
in article <1505@puff.cs.wisc.edu>, avery@puff.cs.wisc.edu (Aaron Avery) says: > Over twice? This must be only in cases of 0ns internal MMU translation. The > '030 can have a 2-clock memory cycle, while the '020 is limited to a 3-clock > memory cycle, so there's this advantage up front. The '020 usually ends up with a 4 clock memory cycle if you've got that MMU in there. But the '030 also has a neat 1-clock burst mode; the CPU can't eat memory at that speed, but it can be used for cache fills and things like that. It does require some external circuitry and certainly some design cleverness, though the 2-clock memory cycle is going to require some design cleverness as well if you aren't willing to spend a small fortune for ridiculously fast memories (at the price of memory chips these days....). > Aaron Avery (avery@puff.cs.wisc.edu) -- Dave Haynie "The B2000 Guy" Commodore-Amiga "The Crew That Never Rests" {ihnp4|uunet|rutgers}!cbmvax!daveh PLINK: D-DAVE H BIX: hazy "I can't relax, 'cause I'm a Boinger!"
avery@puff.cs.wisc.edu (Aaron Avery) (03/29/88)
In article <3540@cbmvax.UUCP>, daveh@cbmvax.UUCP (Dave Haynie) writes: > The '020 usually ends up with a 4 clock memory cycle if you've got that MMU in > there. But the '030 also has a neat 1-clock burst mode; the CPU can't eat > memory at that speed, but it can be used for cache fills and things like that. > It does require some external circuitry and certainly some design cleverness, > though the 2-clock memory cycle is going to require some design cleverness as > well if you aren't willing to spend a small fortune for ridiculously fast > memories (at the price of memory chips these days....). > > Dave Haynie "The B2000 Guy" Commodore-Amiga "The Crew That Never Rests" > {ihnp4|uunet|rutgers}!cbmvax!daveh PLINK: D-DAVE H BIX: hazy True, the '030 has burst mode, and no, it doesn't really require any more external circuitry then normal dynamic ram design stuff. But, that nibble mode memory which supports said burst mode is not as cheap as "standard" DRAM. And I say that if you're designing such a high-end product, you should go all out. You'd want to have at least a 64k external cache running at a 2-clock memory cycle, and nibble mode dynamic ram supporting the faster burst-mode cache fill, to counteract the slower random access speed. You would need to add some extra design cleverness to accomplish the compatibility between the two memories, as the cache won't be able to write those 4 burst-mode bits as fast as they're fetched, and probably some others. I believe that this type of design running at 20 or 25MHz with a 68882 (32-bit interface, of course) would create a very impressive amiga system, don't you? Aaron Avery (avery@puff.cs.wisc.edu) ({seismo,caip,allegra,harvard,rutgers,ihnp4}!uwvax!puff!avery)
mike@ninja.cc.umich.edu (Michael Nowak) (04/01/88)
In article <46958@sun.uucp> cmcmanis@sun.UUCP (Chuck McManis) writes: > >With the FlickerFixer in a 2000 you get both at once. The FlickerFixer >has a multisync monitor output on it and the Amiga output is still available >so you can dedicate that output to the genlock'd video you are sending to >the VCR. > > >--Chuck McManis >uucp: {anywhere}!sun!cmcmanis BIX: cmcmanis ARPAnet: cmcmanis@sun.com >These opinions are my own and no one elses, but you knew that didn't you. This may have been stated earlier but what is the availability of the FlickerFixer and what is the cost? ------------------------------------------------------------------------------ In Real Life: Michael Nowak "Seek truth from facts." Via Internet: mike@ronin.cc.umich.edu - Deng Xiao Ping Via UUCP: uunet!umix!ronin.cc.umich.edu!mike Working for but in no way representing the University of Michigan. ------------------------------------------------------------------------------
cmcmanis%pepper@Sun.COM (Chuck McManis) (04/01/88)
In article <461@mailrus.cc.umich.edu> (Michael Nowak) writes: > This may have been stated earlier but what is the availability of the > FlickerFixer and what is the cost? Well a couple of the dealers around here (N. Cal) have them in stock so I guess availability is 'off the shelf'. I am not sure of the price so I can't comment on it. --Chuck McManis uucp: {anywhere}!sun!cmcmanis BIX: cmcmanis ARPAnet: cmcmanis@sun.com These opinions are my own and no one elses, but you knew that didn't you.