NetNews Usenet Archive 1992 #30

home *** CD-ROM | disk | FTP | other *** search

/ NetNews Usenet Archive 1992 #30 / NN_1992_30.iso / spool / comp / sys / amiga / misc / 18508 < prev next >

Wrap

Internet Message Format | 1992-12-14 | 14.2 KB

Xref: sparky comp.sys.amiga.misc:18508 comp.sys.amiga.advocacy:31283 Path: sparky!uunet!spool.mu.edu!agate!abel From: abel@ocf.berkeley.edu (Abel Chow) Newsgroups: comp.sys.amiga.misc,comp.sys.amiga.advocacy Subject: comments re: WOC-Toronto future of Amiga Date: 14 Dec 1992 13:10:42 GMT Organization: U.C. Berkeley Open Computing Facility Lines: 367 Distribution: world Message-ID: <1gi14iINNo7r@agate.berkeley.edu> NNTP-Posting-Host: 128.32.234.23 OK. I'm all excited now. I'm between finals now so I've got a little time to comment . . . WARNING: This is all speculation, though it's pretty much based on what I read (on the net and otherwise) and what I know about computers. > World of Commodore Toronto 1992 > Lou Eggebrecht tells of the Future [stuff about who Lou is, etc. deleted] > In the future there will be two versions of the chipset: > i) Low End > ii) High End > > Low End: > - Two 100k Transistor Chips Is this using CMOS? Is there a reason that this couldn't be on a single chip? (Maybe it's a limit to the pin count . . . Or perhaps the two chips are actually functionally separate, i.e. video on one and I/O on the other.) What's the level of integration (is that how you say it?), i.e. using 0.6 micron technology (which is currently at the cutting edge)? > - 32 bit DRAM 60ns Page Mode Chip Memory This CHIP RAM would be faster than my FAST RAM (but then I have an A500). Sweet. > - 57 MHz Pixel Clock OK. That means that we can get double the resolution (or the same resolution without the interlace). > - 100% Backwards Compatibility Hm. Better not let those demo-coders in c.s.a.programmer know about this. They have enough trouble looking one generation ahead. Heaven knows if they're able to look two generations ahead. > - 4MB 4Mbps Floppy Controller with Hardware CRC What's the speed right now? What does the 4MB denote? What advantages does this have over generic PC floppy controllers? Personally, I don't think it'll affect too many people since floppies are (at least on non-dirt-cheap machines) not used terribly often with hard drives available. Anyhow, I'd guess this'll be able to read/write HD disks (how about quad density?). > - 8x Memory Bandwidth Increase 2x from 32 bit memory, 2x from page mode memory, and 2x from 2x clock speed, right? > - 2x Blitter Performance ( gets twice as many clocks as on AGA ) OK. 2x from 2x clock speed. > - 800x600x8bit Non-Interlace 72Hz Refresh Rate > - Larger screens at lower Refresh Rates Hm. So, 1280x400x8 (60 Hz) and 1280x512x8 (50 Hz) non-interlaced. Thus, 1280x800x8 (30 Hz), 1280x1024x8 (25 Hz) should be possible interlaced . . . Or how about 1024x768x8 (36 Hz)? (I believe all this can be figured out solely on the speed of the pixel clock.) BTW, 1280x1024 is the top resolution of SGI workstations right now. > - 16 bit True Color mode ( although recent developments with the > completion of the first cycle of chipset design indicate that this > will actually be a 24 bit True Color Mode ) Hm. 16 bit True Color . . . like the Falcon. <grin> 24 bit True Color would give everyone a run for their money. No more HAM mode! > - FIFO serial ports Does this make serial buffer overflows less likely? I'm not familiar with FIFOs (though I know what they are, I can only guess at its purpose). > - increased Chip Ram ( 8Mb ) Yay! Happy day. At least now, the chipset could handle double buffering a 1280x1024x8 (interlaced) screen. Personally, with 1 MB of CHIP RAM, I've never felt the crunch for it. I guess I don't run graphics intensive stuff. But I figure with the higher resolution modes, 2 MB could be pretty limited with the image sizes pushing a MB (depending on the mode). > High End: OK. Now for the really meaty stuff . . . <slobber> > - 4 Chips ( 750k transistors each ) Hm. I'll comment on this later. > - 32/64 bit VRAM Chip Memory ( could be DRAM for mid range machines ) Hm. 64 bit 30 ns memories VRAM (I don't know how fast they really go. This is just a guess.)? That'd give 2x 2x = 4x the performance of the low end chip set. Of course, that would be mighty expensive. Would the CPU still be able to write to VRAM? > - 57/114 MHz pixel clock > - Chunky Pixel Mode as well as Bit Planes Modes ( Blitter works with > both ) Hm. Best of both worlds. > - Special CD-ROM Port ( 4 Mbps Very High Speed Serial Port ) Hm. An audio CD runs as 172 KB/s. Any plans for a double-speed CD-ROM like the ones that are appearing for the PCs? > - Built in Frame Grabber ( Cool! ) Wow. This MUST be high end. > - 1280 x 1024 24 bit color 72 Hz Refresh Rate Hm. That's very impressive. Nice, crisp, stable image. SGIs run at 60 Hz at that resolution (but I don't know if they're able to go faster). > - 16 bit 8 channel 100KHz sampling rate Audio ports Better than CD quality sound!!! Wow. Two output channels? This could be the workings of a digital audio mixing studio (a la AudioMedia's sound board for the Mac). All that's needed are a bunch of XLR inputs (to a sampler), fast I/O (which the SCSI should take care of), and digital outputs (is this expensive to do?). > - New 'On-Demand' DMA Architecture for Balanced DMA Usage I'm completely clueless on this point. Somebody enlighten me, please. > - All Clocks are Independent and Each Section is Asynchronous to Each > Other OK. Good for the long-term. > - 12 to 20 x Memory Bandwidth Increase ( mostly from VRAM ) Hm. So this is 50% to 150% faster than the low end chip set. I guess part of it is 2x clock speed (which the slower memory can't keep up with). And then going to 64 bit memory would be another 2x speedup (which again, the memory can't keep up with). > - 32 bit blitter ( instead of word blitter ) 8x Performance Increase 4x from 4x clock speed (WILD guess). 2x from 32 bits. Of course, since we're in the neighborhood, let's compare with SGI. PowerVision has a 20 million pixels/sec. DMA rate (whatever that means) and can draw 200 million pixels/sec. pixel fill rate. I think there's quite a ways to go (at least at the animation side of things). I wonder if there are any plans for more sophisticated graphics operations (i.e. 3D transformations, shading, lighting, etc.). > - 24 bit True Color Mode Watch out SGI! > - Video Upgrade module ( You can add more chips for parallel processing > chip set i.e. multiple blitters, and Higher Resolution Display Modes ) I wonder how the blitters work in parallel. One per bitplane? Anyhow, this definitely sounds like graphics workstation territory. > - Hardware Graphics Decompression Modes JPEG. MPEG. What else? As part of chipset or off-the-shelf? > - ECS & AGA compatibility This'll be moot WHEN RTG comes (and soon, I hope). > - 32 bit Processor Independent Processor Bus ( hinted at RISC by > calling the bus 'RISC ready' ) Whatever. > Future releases of AmigaDOS: > > - 3.1 Device Independant Network Support Finally. Once there is OS support, 3rd parties with fill in the niche. Then Amigas will be networked like Macs are (well, maybe not THAT easily, but a lot easier and cheaper than right now). > File and Printer Sharing Goes with above. That probably means that a there needs to be a print spooler (it would be very lame if it had only locking). > DSP Support Does anyone know anything about the DSP? > - 4.0 Retargetable Graphics OK. What everyone's been waiting for. > Full Support for Postscript Printers Hm. What kind of support? Fonts? > He also fielded many good questions, and he indicated: > > - The only reason that the A600 and A1200 do not have the dual speed floppies > is that the current model is simply too big. There is lower profile versions > on the way but there are not here now and will not be for a while. > > - There are no plans for virtual memory. > > - Commodore will release a series of Quad-Syncing monitors in February to > replace their current series. Hm. 50, 60, and 72 Hz and what else? (Yes, I know, but I'm not sure how to do the conversions.) > - New Bridgeboards will be on the way, including new software VGA support and a > new bridgeboard for the A1200. Alright. Sounds good. > - AGA will be put into all models of the Amiga, including the CDTV. Also CDTV > technology will be brought to the entire Amiga line. Hm. So this is what they mean by low-end (the AGA chip set, that is). Quite nice. > - Full motion video support is well on its way. The feature is here now, but > the development systems for this technology is not. > > - Prime goal with the CDTV is to cost reduce it and then enhance it. > > - A new design philosophy is to design all hardware with modules to allow > upgradability. Good plan. > - A main goal of his is to use the current state of the art development > systems and design tools to enable Commodore to reduce their design times. > Much of it is already in place and the AGA chipset is the first result of > that process. Instead of being last to bring something out, Commodore is > going to be first or second. Ah, but aren't most companies going to try to take advantage of the development of CAD/CAE tools? Anyways, the results can be very dramatic. At Intel, the productivity of the hardware designers grows an order of magnitude each generation of the software. > - DSP technology will be available for the A4000 in summer. DSP will feature > the AT&T 32000 series DSP and a 68040 processor module for the A4000. A > Personality board will attach to this board and will allow the DSP to perform > useful functions. Most of the DSP Personality boards will be produced by > third parties. There will be extensive information on this product at the > next DevCon. Once again, what is the AT&T 32000? > - The A3090 will be shipping at the end of January and will feature a 10MB/s > transfer rate as well as WIDE and FAST SCSI. <jaw drops to ground> > - A new, higher performance Ethernet card will be available soon, after > the release of 3.1 and the new TCP/IP and Novell software. > > - The chipset is still too power hungry for a laptop. Maybe there will be one > when there is a fully CMOS chipset. It'll be a very long while. > - No hardware memory protection plans for Motorola Architecture ( the way he > said it implied that it might be on other architectures. ) > > - Retargetablity is a major goal, and Audio will merge with DSP ( this didn't > make sense to me since audio is built into the chip set. ) Well, the audio is still accessed through the library. Instead of going to Paula, it'd go to the DSP. Oh, boy, FFTs! > - Processor Upgrade cards for the A1200 will be left for 3rd Parties. Fair enough. The engineers should work on the high end stuff. <grin> > - They are going to be replacing the 2000 very soon. The only reason they > still make the 2000 is so that they don't abandon the only machine the Video > Toaster runs on out of the box. The NewTek people are very difficult to > work with and are resistant to change. <sigh> Tough to negotiate either way. > - Some early 4000's shipped with the fan installed backwards and this caused > many overheating problems. I believe he said that if you had one, you can > have it fixed on your warrantee. That sucks! (Sorry, Skip.) > - He could not give any release dates for machines with the low and high end > chip sets. He did say it would be more than a year. The high end chipset > development is father along than the low end. Both have completed a first > design cycle. The results of this cycle are better than expected and they > are ahead of their development plan. Hm. I guess it'd be too late for an educational discount, then. > One thing that was heavily implied was that the Amiga will be migrated > over to a RISC CPU in the future. While he did metion that the exact > CPU has not been decided, the high-end models are being designed to > accomodate a RISC CPU. I overheard him talking about being excited that > a RISC Amiga will be able to run not only UNIX, but Windows NT. Alpha? MIPS? PA-RISC (ha, fat chance!)? Sparc? PA-RISC is the king of the hill but I don't think HP'd let other's make use of it. Sparc has a software base, but it's lagging in performance. I don't know much about the Alpha but it's fast (and disappates lots of heat). The MIPS processor would be ideal if the GL library were ported to take advantage of the new chip sets. Right now, SGI machines which use the MIPS processor have the most graphics applications out there. Such a move would be very wise, in one fell swoop taking advantage of the extensive software collection built up for it. Ideas? > The timetables for the new stuff were not made clear. He did mention > that the DSP CPU module for the 4000 would be out in the summer. This > implies that 3.1 would be out at the same time. Futher, since 3.1 > includes device independant networking built into the operating system, > the new Ethernet hardware and software should be out shortly after that. > Other products, like the 3090 disk controller and the new monitors, will > be released during the next few months. Hm. How do I earn lots of money in a short period of time??? > The crowd attending the two seminars that Lou gave seemed overwhelmed at > the features of the future chipset. The loudest cheers came with the > mention of the larger chip ram feature. Hm. . . . er. <no comment> > I am very excited about the future of the Amiga. The new features are > nice, but the best part is that, clearly, Commodore has a plan. This plan > is not just a short term plan, but a visonary plan to create a series of > Amigas that will meet future market demands. This plan would not be > reassuring if it was persented by anyone other that Mr. Eggesbrecht. He > is very capable and has had much success in the industry. Me too. It's got me excited about it too. Somehow, this seems different than the past years when we've been disappointed time after time. > DISCLAMIER: > > I don't want to start any dangerous rumors here. Don't take the above as the > absolute truth, since my version of what he said and what he actually said > (or meant) might be different. Neither do I, but I probably just opened up a can of worms. Yeah, what he said. > Shawn Switenky > Shawn_E_Switenky@engr.usask.ca Abel Chow (abel@ocf.berkeley.edu) -- ==E =D--Abe (abel@ocf.berkeley.edu)