dtynan@sultra.UUCP (Der Tynan) (11/04/88)
OK, so I've been following all these postings about so-and-so who makes an
"ultra-cheap" 32K experimental system. Well, so far, 'ultra-cheap' has been
open to varied interpretation. Most of this stuff (in fact ALL of this stuff)
is out of my price-range. Yes, I *do* want it buttered on both sides. I
want high-performance, AND I want low-cost. So, I figure the best way to
get a 'quick-and-dirty' 32K system, is if someone out there (read National)
has a simple system, for which they're willing to sell just the bare PC
board. I mean, when you take manufacturing costs out of the equation, it
should be possible to produce a simple 6MHz, 512K RAM PC board, with a UART
and space for the TDS PROM's for less than $100. Batteries not included.
In reality, it would be nice to find something to plug into my IBM-PC (make
it into a *real* system), which didn't cost $800.00 (and that's the DISCOUNT
price!). As far as an OS is concerned, we can all sit around and wait for
GNU... Or maybe use MINIX... Who knows? I'd just like to dive into the 32K
world, AND get away from that disgusting Intel architecture in one fell swoop.
Or, am I just dreaming again... Comments anyone???
- Der
PS;
The volume is beginning to drop off again... I don't want to have to get
nasty...
--
Reply: dtynan@sultra.UUCP (Der Tynan @ Tynan Computers)
{mips,pyramid}!sultra!dtynan
Cast a cold eye on life, on death. Horseman, pass by... [WBY]grenley@nsc.nsc.com (George Grenley) (11/05/88)
In article <2613@sultra.UUCP> dtynan@sultra.UUCP (Der Tynan) writes: >OK, so I've been following all these postings about so-and-so who makes an >"ultra-cheap" 32K experimental system. Well, so far, 'ultra-cheap' has been >open to varied interpretation. I still maintain that if cheap is your primary goal, you can't beat the 532DK, at a whopping $532. Yes, half a k-buck is not trivial, but you have something which, with work, can be a real OS9/Minix/GNU/whatever system. Starting with 016/032 based hardware means, when you're all done, it will be out of date. Would you build an 8088 based box? No. > I mean, when you take manufacturing costs out of the equation, it >should be possible to produce a simple 6MHz, 512K RAM PC board, with a UART >and space for the TDS PROM's for less than $100. Batteries not included. Der, the realities of corporate bookkeeping are such that we can't do ANYTHING in low volume for $100. Of the $7800 list price of the VME532, for example, the manufacturing labor is significantly less than 10%. So, if you want to have fun building a machine you'll find useful when you're done, buy a 532DK. Regards, George Grenley
gbs@stratus.UUCP (George B. Smith) (11/05/88)
In article <2613@sultra.UUCP> dtynan@sultra.UUCP (Der Tynan) writes: >OK, so I've been following all these postings about so-and-so who makes an >"ultra-cheap" 32K experimental system.... >...So, I figure the best way to >get a 'quick-and-dirty' 32K system, is if someone out there (read National) >has a simple system, for which they're willing to sell just the bare PC >board. I think there is something that you should understand about National. I am not saying the following is good or bad, it's just the way it is. National is not interested in supplying low cost solutions for customers that are only going to buy *one* chip set. They have geared their whole 32k organization to look for, cater to, communicate with, ship to, etc, etc, very large companies. They can deal very well with another large company but don't do well at all with an individual. Once you realize this inherent characteristic of National a lot of their actions make more sense (not good sense, just more sense :-)). >I mean, when you take manufacturing costs out of the equation, it >should be possible to produce a simple 6MHz, 512K RAM PC board, with a UART >and space for the TDS PROM's for less than $100. In large quantities, you are right, it can be done, minus RAM costs which are variable right now. The point is, National doesn't think they would sell very many of them. And I really want to emphasize that this is their perception. Whether they would sell a million or not is not the point. They don't think they will so they won't do it. As an aside, you can forget the TDS PROM's. As one who has been inside the TDS code, you would do well to erase any PROM which has TDS in it (sorry Rich :-)). TDS, which stands for the Tiny Developement System, is the largest piece of 32k code which can stand as the best example of how *not* to write assembly code for the 32k. Also, it accepts only a subset of the old NSX format assembly language, the editor is *very* limited, the assembler is unexplainably slow, on and on. The engineer who last worked on TDS for the 532 designer kit was driven up the wall trying to work with the code. They said on *many* occasions that it would have been better to start from scratch than to use TDS but they had to follow orders. >In reality, it would be nice to find something to plug into my IBM-PC (make >it into a *real* system), which didn't cost $800.00 (and that's the DISCOUNT >price!). I'm afraid that a 6 MHz 016 can not really be used as the basis for "a *real* system" at this date and time. Also, there are no really cheap ways to get into the 32k world. Believe me I have looked high and low since I believed that someone if not National should offer such a solution. I must confess that I found no satisfactory answers. The Ziaz offer recently posted seemed as good as I have seen altho I have no direct experience with them. I have talked to people who I respect that said they were a good outfit, tho. >...get away from that disgusting Intel architecture in one fell swoop. Now this sentiment I can understand and agree with entirely :-). >dtynan@sultra.UUCP (Der Tynan @ Tynan Computers) George B. Smith disclaimer: these opinions are mine (used to be gbs@nsc) and mine alone. Stratus Computer, Inc gbs@stratus
mlewis@unocss.UUCP (Marcus S. Lewis) (11/06/88)
In article <2613@sultra.UUCP>, dtynan@sultra.UUCP (Der Tynan) writes: > so far, 'ultra-cheap' has been > open to varied interpretation. Most of this stuff (in fact ALL of this stuff) > is out of my price-range. Yes, I *do* want it buttered on both sides. > board. I have a CompuPro CPU32016. I have been playing with S-100 hardware for over 10 years now, and have collected a few dozen odd boards and a couple of odd boxes. This for me is cheap. I spent $95 for my CPU32016. The rest of it I already had "in stock". No, the system isn't running. Yes, I have all the parts I need. I don't have the time I need. > As far as an OS is concerned, we can all sit around and wait for > GNU... Or maybe use MINIX... Who knows? I'd just like to dive into the 32K > world, AND get away from that disgusting Intel architecture in one fell swoop. > Or, am I just dreaming again... Comments anyone??? > - Der So get off yer duff and add your input to the OS group I have been nattering about in all my postings. Or do you just want to bellyache about what other people think an OS oughta do? If you have a Z80 system running, you can have all the development tools you need to get started. Not a Sun Unix workstation, but an assembler that screams and has very few nasty surprises. If you have an MS-DOS machine, you can run the Z80MU program and have more of the same. We also have C compilers, and yes, a choice of assemblers. If that's not enough, write yer own and contribute your expertise. We need help, not criticism. Marc Lewis
mlewis@unocss.UUCP (Marcus S. Lewis) (11/06/88)
In article <7532@nsc.nsc.com>, grenley@nsc.nsc.com (George Grenley) writes: > Starting with 016/032 based hardware means, when you're all done, it will be > out of date. > > Would you build an 8088 based box? No. > If that was the cheapest way to get my feet wet, you betcher tootsies. Since NSC thoughtfully has "committed" to retaining *binary* compatibility across the 32K family, this is a loaded, and therefore unfair, question. I have a 32016, and I intend to make it run efficiently, and learn the architecture while writing the OS to end all OS's. Then, I plan to have all of my kids through high school (not too long now!), and making a decent living. Right now, $100 put a real dent in my paycheck, and there's no way I can handle $532, plus a PS, a box, a decent set of disk drives, and a purchased OS. So, piecemeal, I am building my S-100 system. I am using my non-PC MS-DOS machine to develop code, and spending what free time I have designing code, and incidentally learning in depth OS design, as well as paractical compiler design. Learning was what I bought my first machine for, and even though I am finishing a MS in Applied Math, I don't think I want to stop learning. When I finish my system, I will be in a better position to decide if I *really* need a '532 or '732 or '932, and its associated vastly increased performance. As it is, an 8 MHz 32016 will run circles around my 7.5 MHz V20/8088, and any CP/M system I happen to have that still runs. Most of us "hobbyists", even pros, don't need the power of a Cray4, we just think we do. That's the nature of these expensive toys. Marc Lewis
gbs@stratus.UUCP (George B. Smith) (11/06/88)
In article <7532@nsc.nsc.com> grenley@nsc.nsc.com.UUCP (George Grenley) writes: >In article <2613@sultra.UUCP> dtynan@sultra.UUCP (Der Tynan) writes: >>OK, so I've been following all these postings about so-and-so who makes an >>"ultra-cheap" 32K experimental system. Well, so far, 'ultra-cheap' has been >>open to varied interpretation. > >I still maintain that if cheap is your primary goal, you can't beat the 532DK, >at a whopping $532. Yes, half a k-buck is not trivial, but you have >something which, with work, can be a real OS9/Minix/GNU/whatever system. I suggest that the 532DK, otherwise known as the 532 Designer Kit, would *not* be an appropriate starting point for a system. There is no bus, there is only 128k of static RAM with no provision to expand it, there is no provision for adding a mass storage device, i.e. disk of any kind, there is no provision to add networking, etc, etc. The 532DK is just what George Grenley stated before, a demo board which proves that it is possible to wire a 532 up to other chips and run small benchmarks generated by expensive 32k software tools which must run on expensive host systems. There is a "wire wrap area" on the board which George must be referring to when he says that "with work" the 532DK "can be a real OS9/Minix/GNU/whateve system". I would say a *lot* of work that would be of questionable value. Now the CG16 group is working on something that would be a good starting point for a 32k PC system. Hey Dave, yeah, Dave Rand, care to share some details with the net? >Der, the realities of corporate bookkeeping are such that we can't do >ANYTHING in low volume for $100. National does not want to cater to the low cost , low volume market. Now my question is of the chicken or the egg variety, is it because they in fact *can't* do it or they just don't want to? Let's put it another way, if they thought they *could* do it, would they find a way to get it done? >Regards, >George Grenley Regards, the other George George B. Smith Disclaimer: don't hold anyone else Stratus Computer, Inc. accountable but me. gbs@stratus
night@pawl17.pawl.rpi.edu (Trip Martin) (11/07/88)
In article <1039@stratus.UUCP> gbs@stratus.UUCP (George B. Smith) writes: >I suggest that the 532DK, otherwise known as the 532 Designer Kit, would *not* >be an appropriate starting point for a system. There is no bus, there is only >128k of static RAM with no provision to expand it, there is no provision for >adding a mass storage device, i.e. disk of any kind, there is no provision >to add networking, etc, etc. The 532DK is just what George Grenley stated >before, a demo board which proves that it is possible to wire a 532 up to >other chips and run small benchmarks generated by expensive 32k software >tools which must run on expensive host systems. There is a "wire wrap area" >on the board which George must be referring to when he says that "with work" >the 532DK "can be a real OS9/Minix/GNU/whateve system". I would say a *lot* >of work that would be of questionable value. It depends on the individual. For a computer engineering student like me, it's a great deal. For less than the cost of a cheap PC clone, I get a '532 (probably cheaper than if I bought the '532 by itself, although I haven't been able to get prices yet), plus the additional hardware to get a system up pretty quickly. I can also learn from it as an example of a high-speed system, even if it isn't necessarily production quality. Eventually, I could use the '532 in a system of my own design. Granted, I'm talking about a lot of work, but it's also a lot of experience. As for software, Gnu supports the 32000 series. Besides, I don't think it would be too difficult to write a simple assembler for experimenting with the 32532. I would say your comments apply to companies thinking of using the '532, but there are quite a few people who could get a lot of of the 532DK. >National does not want to cater to the low cost , low volume market. >Now my question is of the chicken or the egg variety, is it because >they in fact *can't* do it or they just don't want to? Let's put it >another way, if they thought they *could* do it, would they find a way >to get it done? I don't blame National for not catering to the low-cost, low-volume market, since by definition, there is little profit to be gained by it. If you were National, would you want to spend lots of resources on a minimal profit-margin market? -- Trip Martin night@pawl.rpi.edu userffs7@rpitsmts.bitnet Trip Martin night@pawl.rpi.edu userffs7@rpitsmts.bitnet
dlr@daver.UUCP (Dave Rand) (11/07/88)
In article <1039@stratus.UUCP> gbs@stratus.UUCP (George B. Smith) writes: >Now the CG16 group is working on something that would be a good starting >point for a 32k PC system. Hey Dave, yeah, Dave Rand, care to share some >details with the net? > SSHHH! It isn't released yet. I'll post details as I can. -- Dave Rand {pyramid|hoptoad|sun|vsi1}!daver!dlr
grenley@nsc.nsc.com (George Grenley) (11/08/88)
I'm happy to see Mr. Smith helping out here. As a former NSC employee, he has detailed knowledge of Life Here in 32000 Land. In article <1039@stratus.UUCP> gbs@stratus.UUCP (George B. Smith) writes: >In article <7532@nsc.nsc.com> grenley@nsc.nsc.com.UUCP (George Grenley) writes: >>In article <2613@sultra.UUCP> dtynan@sultra.UUCP (Der Tynan) writes: >>>OK, so I've been following all these postings about so-and-so who makes an >>>"ultra-cheap" 32K experimental system. Well, so far, 'ultra-cheap' has been >>>open to varied interpretation. (GRG talking:) >>I still maintain that if cheap is your primary goal, you can't beat the 532DK, >>at a whopping $532. Yes, half a k-buck is not trivial, but you have >>something which, with work, can be a real OS9/Minix/GNU/whatever system. (GBS talking:) >I suggest that the 532DK, otherwise known as the 532 Designer Kit, would *not* >be an appropriate starting point for a system. There is no bus, there is only >128k of static RAM with no provision to expand it, there is no provision for >adding a mass storage device, i.e. disk of any kind, there is no provision >to add networking, etc, etc. The 532DK is just what George Grenley stated >before, a demo board which proves that it is possible to wire a 532 up to >other chips and run small benchmarks generated by expensive 32k software >tools which must run on expensive host systems. There is a "wire wrap area" >on the board which George must be referring to when he says that "with work" >the 532DK "can be a real OS9/Minix/GNU/whateve system". I would say a *lot* >of work that would be of questionable value. I agree. I said a lot of work would be required. However, for a hobbyist on a budget, it would not be too difficult to add a DRAM controller and a couple of meg of DRAM; and also add a SCSI port. Here at the ol' factory we have a list of app notes we're working on; these subjects are to be covered, as quick as we can get to them. (By the way, while we cannot commit to anything, we do sometimes work with outside engineers to write such app notes.) This would make a pretty good single or small multi user system. Not a NeXT or anything, but the cost is reasonable. >Now the CG16 group is working on something that would be a good starting >point for a 32k PC system. Hey Dave, yeah, Dave Rand, care to share some >details with the net? I do not believe this is a good alternative, for several reasons. First, the 32CG16 has no MMU capabilities, so any Unix type box is out. Second, while the BitBlt capabilities of the CG16 make it quite fast at bit diddling, in terms of general performance, it is more like a 15 mhz 68000 - not that fast compared to other CPUs on the market. Although I haven't benchmarked it, I would guess that a fast 286-clone machine running Microport Unix (or similar) would be at least as fast, maybe faster. Why build a slow machine? Most of the cost of any Unix-capable machine is going to be memory (32 1 meg RAMS = $500-$1000) and disk (40 meg = $500). So, why not hook it up to a GREAT CPU? >>Der, the realities of corporate bookkeeping are such that we can't do >>ANYTHING in low volume for $100. >National does not want to cater to the low cost , low volume market. >Now my question is of the chicken or the egg variety, is it because >they in fact *can't* do it or they just don't want to? Let's put it >another way, if they thought they *could* do it, would they find a way >to get it done? As my esteemed former collegue knows, wanting to do something and doing it are different. NSC, with the overhead of a billion dollar semi plant capacity, simply cannot afford to do business "retail". If some other company wanted to do a PC class machine, we would certainly be interested in seeing them be successful, but frankly I doubt that you can knock off either the DOS camp or the Apple camp. Sun might; but they've got a running start. NeXT might; they've got Steve Jobs. But I don't believe the low end of the computer market can support that many families - the three that exist now (DOS/Intel, Mac/68020, Unix/68020-sparc) are enough, if not too much. >>Regards, >>George Grenley >Regards, >the other George >George B. Smith Disclaimer: don't hold anyone else >Stratus Computer, Inc. accountable but me. >gbs@stratus Best Regards, the first George Nat'l Semi
grenley@nsc.nsc.com (George Grenley) (11/08/88)
TRUTH IN ADVERTISING: In article <7012@daver.UUCP> dlr@daver.UUCP (Dave Rand) writes: >In article <1039@stratus.UUCP> gbs@stratus.UUCP (George B. Smith) writes: >>Now the CG16 group is working on something that would be a good starting >>point for a 32k PC system. Hey Dave, yeah, Dave Rand, care to share some >>details with the net? >SSHHH! It isn't released yet. I'll post details as I can. >Dave Rand >{pyramid|hoptoad|sun|vsi1}!daver!dlr Just in case everyone doesn't know, I should mention that Dave Rand is currently working for National Semiconductor, in the 32000 group. Although, like all of us in this group, he is NOT an official spokesman of any sort. Just a grunt engineer, like me, Ron Guilmette, and a few others who post here from time to time. Although the path name, signature, Organization, etc, don't indicate this, it is the case. Regards, George Grenley NSC
dtynan@sultra.UUCP (Der Tynan) (11/08/88)
In article <7532@nsc.nsc.com>, grenley@nsc.nsc.com (George Grenley) writes: > In article <2613@sultra.UUCP> I wrote: > > > >OK, so I've been following all these postings about so-and-so who makes an > >"ultra-cheap" 32K experimental system. Well, so far, 'ultra-cheap' has been > >open to varied interpretation. > > I still maintain that if cheap is your primary goal, you can't beat the 532DK, > at a whopping $532. Yes, half a k-buck is not trivial, but you have > something which, with work, can be a real OS9/Minix/GNU/whatever system. > Starting with 016/032 based hardware means, when you're all done, it will be > out of date. > > Regards, > George Grenley I thought the 532DK was just a designer kit (all five chips + TDS proms). If you want to actually do something with it, you've got to spend months, nay, years, wirewrapping :-)... What exactly IS the 532DK? Another (hairbrained) idea would be to produce a '332 (or even '532) mother- board for the IBM-PC. I mean, you can pick up individual clone parts (PSU, case etc) real cheap these days. So, you have a motherboard which has up to 8Mb of SIMMS, and a lookalike PC-AT bus, so you can put off the shelf AT-type plug-ins into the box. The National chips are IDEAL for this, because they allow you to dynamically 'collapse' the bus. In this way, the internal bus to the memory chips would be 32-bits wide, but would shrink down to 16-bits, to access stuff of the PC-bus. I'd do this myself, if I had the time and energy. Talk about a wolf in sheeps clothing... - Der -- dtynan@Tynan.COM (Dermot Tynan @ Tynan Computers) {apple,mips,pyramid,uunet}!zorba.Tynan.COM!dtynan --- God invented alcohol to keep the Irish from taking over the planet ---
mlewis@unocss.UUCP (Marcus S. Lewis) (11/08/88)
Preface: If my previous response to this actually made it, I don't know, and I apologize for cluttering things up. I just can't let this go by: In article <7532@nsc.nsc.com>, grenley@nsc.nsc.com (George Grenley) writes: > In article <2613@sultra.UUCP> dtynan@sultra.UUCP (Der Tynan) writes: > >Well, so far, 'ultra-cheap' has been > >open to varied interpretation. > > ... if cheap is your primary goal, you can't beat the 532DK, > at a whopping $532. yup. Add oh, $250 for a decent box and PS, minimum $300 for a terminal, let's guesstimate $800 for a relatively capacious, fast disk drive, and about $1K for an OS that makes it all go. Cheap at any price. > Starting with 016/032 based hardware means, when you're all done, it will be > out of date. > > Would you build an 8088 based box? No. If that's what I can afford, you betcha. Prove to me that a hobbyist needs 15 MIPs, whatever the heck that means, to tinker with. If you are marketing a commercial cruncher, with 16M color palette, etc, ad expensive nauseum, you need that power, but when you are paying your own bills, the 32016 is quite adequate, Thank you. This is especially true when you remember that NSC thoughtfully *committed* to binary compatibility across the whole family. So you build the '016, get it running the Hacker's Dream OS, then find out, maybe you do need that extra oomph, then pop for the $2K for the bigger machine. > So, if you want to have fun building a machine you'll find useful when > you're done, buy a 532DK. Sorry George, I have an original Poly-88 (1.8 MHz 8080) that I still use because it has a BASIC interpreter with 36-digit precision. Slow it may be, but deny it is useful, especially for math classes. Besides I have source for the OS. Marc Lewis
josef@ugun21.UUCP (11/08/88)
In one of the previous responses Trip Martin writes: >If you were National, would you want to spend lots of resources on >a minimal profit-margin market? I agree with You that You couldn't rip lots of $$$$ off people when all they buy is one or two items. BUT ... How many people have thought: "Gee, I would love to take a look at the 32k series, but aren't these buggers expensive?" If the Z80 would still cost $100, no decent CP/M systems would be around, if PCs would still cost $10,000, what would You do? Buy one, write lots of interesting software for it? Why is U**X used by so many universities? Because they paid much money or because they got it almost for free?? In my humble opinion, the abovementioned policy works counterproductive. Josef Moellers paper mail: e-mail: c/o Nixdorf Computer AG USA: uunet!linus!nixbur!nixpbe!mollers.pad Abt. EG-3 !USA: mcvax!unido!nixpbe!mollers.pad Unterer Frankfurter Weg D-4790 Paderborn tel.: (+49) 5251 104691 Standard disclaimer: Blablabla opinion blablabla employer blablabla! /* End of text from ugun21:comp.sys.nsc.3 */
vixie@decwrl.dec.com (Paul Vixie) (11/08/88)
In article <477@unocss.UUCP> mlewis@unocss.UUCP (Marcus S. Lewis) writes:
# I have a CompuPro CPU32016. [...] I spent $95 for my CPU32016. The rest of it
# I already had "in stock". No, the system isn't running. Yes, I have all the
# parts I need. I don't have the time I need.
I know what that's like. A few years back I bought a CPU32016 and all the
junk needed to make it work, disk controllers and serial ports and such. I
wrote an assembler in C (using a Mac of all things). That's it -- end of
project. I read comp.sys.nsc.32k in hopes that someone will come up with
something I can use the darned thing for (besides a doorstop / footrest).
--
Paul Vixie
Work: vixie@decwrl.dec.com decwrl!vixie +1 415 853 6600
Play: paul@vixie.sf.ca.us vixie!paul +1 415 864 7013ward@cfa.harvard.EDU (Steve Ward) (11/09/88)
... There does seem to be active interest in a low-cost approach to implementing a 32K machine of some sort. The idea of making it compatible with PC hardware is a great way to keep the cost down. There is all that chassis and peripheral and interface hardware compatible with the PC/XT bus to be used, and many people probably have such hardware. It seems to me that the only to get such stuff is to make it ourselves. We need volunteers to contribute to such an effort. Here is what I can do: Use a CAE system to generate schematics and printed circuit board layout. This will be made available on MSDOS disks in HPGL file format for schematics and layout and also Gerber photoplot file format for the artwork. Probably, can also supply files in Epson/IBM dot matrix format, though this is not useful for the printed circuit board artwork except for documentation. I can also help with the cpu board circuit design. All this will be distributed public domain to whoever wants it. Volunteers needed for the following: Help with design specs (which 32K chips, also specific design help) and circuit design -- National guys sending applications notes and chip spec/data and programming docs would be a great start Can anyone arrange Gerber photoplotting free or cheap? This is not very expensive but costs real bucks ($100 - $200) How about making the PCB's? The above two manuf. steps may have to be arranged by volunteers contributing x bucks each and sending it out for fab. I can coordinate this activity. Software programming volunteers -- need debug monitor, assembler, etc. Issues need to be discussed: I think there should either be two versions, one with MMU and one without MMU, or a single version that will work either way. Real-time users may not need/want MMU and it is not needed for most of the low cost OS's. Accomodation for FPU is mandatory. These are two examples to get the ball rolling if not the ire up. Anyone volunteering or caring to criticize constructively please send me email and I will summarize. For the record I have no 32K use or design experience as I have been and continue to be a user of Moto parts, but am interested in the 32K parts. For that matter, the same thing can be done using Moto parts and I would be glad to make Schems and Artwork for both. Steven Ward ward@cfa.harvard.edu (617)495-7392
grenley@nsc.nsc.com (George Grenley) (11/09/88)
In article <2621@sultra.UUCP> dtynan@sultra.UUCP (Der Tynan) writes: >In article <7532@nsc.nsc.com>, grenley@nsc.nsc.com (George Grenley) writes: >> I still maintain that if cheap is your primary goal, you can't beat the 532DK >> at a whopping $532. Yes, half a k-buck is not trivial, but you have >> something which, with work, can be a real OS9/Minix/GNU/whatever system. >I thought the 532DK was just a designer kit (all five chips + TDS proms). If >you want to actually do something with it, you've got to spend months, nay, >years, wirewrapping :-)... What exactly IS the 532DK? The 532DK includes a PCB. You can build it yourself, then add DRAM (your design), and a disk interface (your design) and create a complete system. As we supply it, you can write ASM and debug it, and so you have enough to bootstrap yourself up. It's a good start, for $532 Regards, George Grenley NSC
curry@nsc.nsc.com (Ray Curry) (11/09/88)
In article <7619@nsc.nsc.com> grenley@nsc.nsc.com.UUCP (George Grenley) writes: >>years, wirewrapping :-)... What exactly IS the 532DK? > >The 532DK includes a PCB. You can build it yourself, then add DRAM (your >design), and a disk interface (your design) and create a complete >system. Oops, George. Almost correct there. The RAM used is actually static, 64K X4. You can use 25 NS versions for 0 wait state operation or 45 ns for 1, or slower for more waits. The kit includes the NS32532, sockets, the one and only PAL, plus the PC board of course. The builder must supply some simple TTL and the RAM chips.
grenley@nsc.nsc.com (George Grenley) (11/09/88)
We'll get this right yet.... In article <7626@nsc.nsc.com> curry@nsc.nsc.com.UUCP (Ray Curry) writes: >In article <7619@nsc.nsc.com> grenley@nsc.nsc.com.UUCP (George Grenley) writes: >>The 532DK includes a PCB. You can build it yourself, then add DRAM (your >>design), and a disk interface (your design) and create a complete >>system. >Oops, George. Almost correct there. The RAM used is actually static, 64K X4. >You can use 25 NS versions for 0 wait state operation or 45 ns for 1, or slower >for more waits. The kit includes the NS32532, sockets, the one and only PAL, >plus the PC board of course. The builder must supply some simple TTL and the >RAM chips. Ray is correct, but I meant something else, namely that you can add DRAM, but only by designing the interface yourself - i.e., other than the w/w area, we have made no particular provision for it. Let me summarize: The DK is a minimum-feature 532, that will allow you to run simple programs. It is expandible, roughly in proportion to your ingenuity. NSC's purpose it offering it was to provide a low cost way for a user to run a small custom benchmark, and generally "get acquainted" with the 532. Professionals seeking an out-of-the-box platform for more serious development should consider the VME based boards offered by NSC and Heurikon. On the other hand, cash-starved college students & hobbyists should think about the DK. Together with some DRAM & a disk drive, you could probably cobble up a system for < $1500 if you shopped around. Happy Hacking! George Grenley NSC
gbs@stratus.UUCP (George B. Smith) (11/09/88)
In article <7594@nsc.nsc.com> grenley@nsc.nsc.com.UUCP (George Grenley) writes: >I said a lot of work would be required. However, for a hobbyist on >a budget, it would not be too difficult to add a DRAM controller and a couple >of meg of DRAM; and also add a SCSI port. Amen to a lot of work! My viewpoint here is that it would be so much work that a blank PC board may be a better starting point. In other words, what advantage does the 532DK give other than the 532 chip? It is in a non-standard physical format, and it has only 128k of static RAM and a serial port. >Here at the ol' factory we have >a list of app notes we're working on; these subjects are to be covered, as >quick as we can get to them. (By the way, while we cannot commit to anything, >we do sometimes work with outside engineers to write such app notes.) I have to say that I am skeptical about these app notes. The staffing level in the 32k group is the issue here. Now the cg16 group has done some *very* good things in the way of app notes as is readily apparent in the new 32k data book. For netlanders, please note that the 32k app group and the cg16 app group are two totally different groups with very different management. >This would make a pretty good single or small multi user system. Not a >NeXT or anything, but the cost is reasonable. As long as you continue to say this I must continue to disagree. >>Now the CG16 group is working on something that would be a good starting >>point for a 32k PC system. > >I do not believe this is a good alternative, for several reasons. First, the >32CG16 has no MMU capabilities, so any Unix type box is out. What about Minix? XINU? They run very nicely on a PC/XT with no mmu. For that "low cost, hobbyiest system" you mention, I think the cg16 is more than capable. It will be a very low cost part, will be supported *very* well, and will be widely available I suspect. I wouldn't brush it off quite yet. Also, the configuration of the cg16 PC will be *quite nice*. >Second, while the >BitBlt capabilities of the CG16 make it quite fast at bit diddling, in terms >of general performance, it is more like a 15 mhz 68000 - not that fast compared >to other CPUs on the market. Although I haven't benchmarked it, I would guess >that a fast 286-clone machine running Microport UNIX (or similar) would be >at least as fast, maybe faster. I am sure that the cg16 people can defend themselves far better than I can. Let me say that at least it will get around the horrible segmentation problems that the 286 UNIX people have. >Why build a slow machine? Most of the cost of any UNIX-capable machine is >going to be memory (32 1 meg RAMS = $500-$1000) and disk (40 meg = $500). >So, why not hook it up to a GREAT CPU? Since we have been discussing low-cost, "hobbyiest" systems, I don't think many of them will be porting actual UNIX to this type of hardware. Besides, I don't think you can beat the Ziaz deal right now anyway. >>Let's put it >>another way, if they thought they *could* do it, would they find a way >>to get it done? > >As my esteemed former collegue knows, wanting to do something and doing it >are different. NSC, with the overhead of a billion dollar semi plant capacity, >simply cannot afford to do business "retail". I am afraid that I *don't* know. That is why I asked the question. Let's look at your half-empty glass a different way. Suppose we looked at it as a half-full glass, i.e. National, with the resources of a $2 billion company, can it do something that small companies can do with far, far fewer resources can do? I think it is a matter of *desire*. If you wanted to produce a neat low cost, useful, fun system, then you probably would find a way. If you *don't* want to do this sort of thing, then it doesn't matter what you have in the way of resources or overhead or what, you probably won't do it. >Best Regards, >the first George >Nat'l Semi George B. Smith Disclaimer: personal opinion only. Stratus Computer, Inc. gbs@stratus
grenley@nsc.nsc.com (George Grenley) (11/09/88)
In article <1264@cfa.cfa.harvard.EDU> ward@cfa.harvard.EDU (Steve Ward) writes: >There does seem to be active interest in a low-cost approach to >implementing a 32K machine of some sort. The idea of making it >compatible with PC hardware is a great way to keep the cost down. >There is all that chassis and peripheral and interface hardware >compatible with the PC/XT bus to be used, and many people probably >have such hardware. >It seems to me that the only to get such stuff is to make it ourselves. Hallelujah! We at NSC applaud you. Please keep me/us posted, we will do what we can to help. You can email to me at nsc!grenley George Grenley
gbs@stratus.UUCP (George B. Smith) (11/09/88)
In article <1669@imagine.PAWL.RPI.EDU> night@pawl17.pawl.rpi.edu (Trip Martin) writes: >In article <1039@stratus.UUCP> gbs@stratus.UUCP (George B. Smith) writes: >>I suggest that the 532DK, otherwise known as the 532 Designer Kit, would *not* >>be an appropriate starting point for a system. > >It depends on the individual. For a computer engineering student like >me, it's a great deal. For less than the cost of a cheap PC clone, I >get a '532 (probably cheaper than if I bought the '532 by itself, >although I haven't been able to get prices yet), plus the additional >hardware to get a system up pretty quickly. For a fairly complete, take it home and plug it in, *XT PC Clone*, I have seen many machines in the *under* $500 range. The 532DK will require the $532 532DK, a terminal, a serial cable, a power supply, and a power cable to have a running system. Quite a bit *more* than a cheap PC clone and you will need a host system to do any serious work with the 532DK. >I can also learn from it >as an example of a high-speed system, even if it isn't necessarily >production quality. You see, that is the problem. The 532DK is *not* a system! Let me state I am not a hardware person but I can tell you that the 532DK will *not* show you how to interface to a *real* memory system. For a system with the 532, memory would consist of a large array of dynamic RAM with a cache between the 532 and memory. The 532DK has a very simple, small static RAM configuration. The 532DK will not show you how to interface to a mass storage system, will not show you how to interface to a graphics system, will not show you how to interface to an ethernet, or to any kind of bus. It will show you how to interface to a serial port though. What I am trying to say is that I don't think you are not getting what you are looking for. >As for software, Gnu supports the 32000 series. Besides, I don't think >it would be too difficult to write a simple assembler for experimenting >with the 32532. I think it is widely agreed that to run any GNU software from the Free Software Foundation requires a 32-bit UNIX system with substantial disk space. Read this as expensive. As for an assembler, there is a "simple" public domain 32k assembler written in C which runs on DOS available from Rick Rodman of the NS32 User Group. It is an improved version of the one that appeared in Dr. Dobb's Journal a couple of years ago. Note that it has a non-standard syntax (i.e. not compatible with either the current National software tools, GNX R3, or the assembler in the TDS ROMs which come with the 532DK), produces only Intel hex files, and does not support the 532 (unless he has upgraded it in the last two months). >I would say your comments apply to companies thinking of using the '532, >but there are quite a few people who could get a lot of of the 532DK. I remain unconvinced. Now the cg16 PC could change my mind. >>National does not want to cater to the low cost , low volume market. > >I don't blame National for not catering to the low-cost, low-volume >market, since by definition, there is little profit to be gained by it. >If you were National, would you want to spend lots of resources on >a minimal profit-margin market? I was National, at least part of it, for a long while. I *did* want to participate in this market. It doesn't show a profit on this quarters bottom line like the MBA's would like to see, but what about the future? How do you cultivate all the *future* and potentially profitable users that could turn into something big someday? What about the "poor, struggling, resource-poor but intelligent, hard working and energetic students" like yourself? What about the 2 to 5 man gargage shops that pop up all over the place in Silicon Valley? What about the technically sophisticated experimenters who are avid followers of experts like Steve Ciarcia from BYTE magazine? All these people are the engineers of large companies of the future who might well be specifying 32k's *if* they get some consideration *now*. And I don't think it would take a "lot of resources" to do this. In many instances that I observed, it would simply mean not excluding them by arbitrary policies. >Trip Martin >night@pawl.rpi.edu >userffs7@rpitsmts.bitnet George B. Smith disclaimer: this is fun Stratus Computer, Inc gbs@stratus
gbs@stratus.UUCP (George B. Smith) (11/09/88)
In article <2621@sultra.UUCP> dtynan@sultra.UUCP (Der Tynan) writes: >I thought the 532DK was just a designer kit (all five chips + TDS proms). If >you want to actually do something with it, you've got to spend months, nay, >years, wirewrapping :-)... What exactly IS the 532DK? While there has not been any actual 532DK's shipped yet, the configuration as of last month was a 532 on a small board with 128k static RAM, one serial port, one parallel port set up as a centronics printer port, and TDS and mon532 in EPROM plus some documentation. Oh, don't let me forget to mention the wire wrap area for all that "hard work" that George Grenley is going to do. :-) >Another (hairbrained) idea would be to produce a '332 (or even '532) mother- >board for the IBM-PC. I mean, you can pick up individual clone parts (PSU, >case etc) real cheap these days. So, you have a motherboard which has up to >8Mb of SIMMS, and a lookalike PC-AT bus, so you can put off the shelf AT-type >plug-ins into the box. Gee, the cg16 group is not going to like you calling there new baby hairbrained. :-) >dtynan@Tynan.COM (Dermot Tynan @ Tynan Computers) George B. Smith disclaimer: the cat is almost out of the bag Stratus Computer, Inc gbs@stratus
cjosta@taux01.UUCP (Jonathan Sweedler) (11/09/88)
Sounds good to me to, but somebody else had the same idea about 7 months ago. You might want to talk to them first. I tried to get some people here interested in the idea at that time, and they said that we could do the following: > 1. Supply them with our version of Unix V.3 + help (if they need). > 2. Meet them and disscuss H/W issues (caches, speed etc.) > 3. Give them the design info of VME532 (I don't know it they need > it, as their design is very cost sensitive) I don't know what happened with this though, or if the other design effort ever got off the ground. I haven't heard anything about the design for about 6 months or so. Anyway, here is a VERY short summary of the original posting (copied, of course, without permission). ------------------------------------------------------------------------- >From: scwilk@sdrc.UUCP (Ken Wilkinson) [also "...!uunet!sdrc!scwilk"] Subject: THE '532 MANIFESTO" (long) Date: 15 Apr 88 18:29:35 GMT Organization: Structural Dynamics Research Corp., Cincinnati The '532 Manifesto (Apologies to R. Stallman) Purpose: Design and build a computer based on the NSC '532 chip set. The computer will use presently available mass-produced componets to keep cost as low as possible. Exploit cheap available finished IBM bus I/O boards, to take advantage of everybody else's volume. Design will be a modified KISS approach (KISS Plus), "Keep It Simple Stupid" but as fast as possible! ------------------------------------------------------------------------- Ken Wilkinson took a poll to find out what people would want to see in a '532 system and who was willing to help. The response wasn't overwhelming, as I recall, but he did get a few volunteers. I can re-post the entire manifesto or mail it to interested people. Hope this helps and good luck (we're rootin' for you!). -- Jonathan Sweedler === National Semiconductor Israel UUCP: ...!{amdahl,hplabs,decwrl}!nsc!taux01!cjosta Domain: cjosta@taux01.nsc.com
josef@ugun21.UUCP (11/09/88)
I would like to pick up Steven Ward's (ward@cfa.harvard.edu) idea.
If somebody would care to coordinate, I would be willing to participate
in the development of the system.
Being a bit further away than "just around the corner" 8-), I would
offer to produce some software:
>Software programming volunteers -- need debug monitor, assembler, etc.
Josef Moellers
paper mail: e-mail:
c/o Nixdorf Computer AG USA: uunet!linus!nixbur!nixpbe!mollers.pad
Abt. EG-3 !USA: mcvax!unido!nixpbe!mollers.pad
Unterer Frankfurter Weg
D-4790 Paderborn
tel.: (+49) 5251 104691
Standard disclaimer: Blablabla opinion blablabla employer blablabla!grenley@nsc.nsc.com (George Grenley) (11/10/88)
For those who may be interested inbuilding from scratch, I will soon be publishing an app note on a fast, relativley low cost DRAM interface for the 532. I invite and encourage NSC engineers to help out anyone who is designing with the 532 as much as they can. In article <930@taux01.UUCP> cjosta@taux01.UUCP (Jonathan Sweedler) writes: >Sounds good to me to, but somebody else had the same idea about 7 >months ago. You might want to talk to them first. I tried to get some >people here interested in the idea at that time, and they said that >we could do the following: >> 1. Supply them with our version of Unix V.3 + help (if they need). Jonathan, our version of Unix is protected by AT&T license terms. Please do not give it to anyone. Even if they have an AT&T license, the transfer must be recorded. Thanks. >> 2. Meet them and disscuss H/W issues (caches, speed etc.) >> 3. Give them the design info of VME532 (I don't know it they need >> it, as their design is very cost sensitive) Good ideas! Let's help out. George Grenley
dtynan@sultra.UUCP (Der Tynan) (11/10/88)
Hmmm. It seems like a kicked over a few hornets nests. The system I have
been envisioning, is neither a 532DK or a cg16. Basically, it would be
a 'blank' printed-ciruit board for the IBM-PC. No components, nothing.
The design would allow for ALL 5 32032 chips (TCU, ICU, FPU, MMU & CPU).
Preferably running at something like 10MHz, but slower is OK, too. The
board would have space for up 8Mb of SIMM's (although it would run with less).
No PROM's, or anything else, just the above. Also, the RAM area could be
addressable from the 8088, using a 64K window. As far as assemblers and things
are concerned, they would reside on the MS-DOS partition of the IBM-PC. The
ultimate target would be to have a second partition for a dedicated OS, such
as MINIX, or a GNU kernel, or whatever. In this environment, the I/O grunt
work would be done by the 8088 (or '286 or whatever). MINIX would fit in
this environment nicely, because it is somewhat 'threaded'. The I/O kernel
could remain as 8088 code (just random thoughts - maybe not feasable). As
far as the comment that an MMU is not required, I beg to differ. As anyone
from the MINIX environment will tell you, one of the big problems with
MINIX, is the fact that an errant program can crash the system. This is NOT
good. In its simplest form, an MMU provides protection against this. As an
aside, it also provides an interesting project for someone to implement VM
under MINIX (Any takers? :-). Also, I'm assuming that people will stuff the
thing with parts from the 32032 Designer Kit, which already includes the MMU,
so why not use it?
As far as manning the project, I'm willing to be the focal point, so that all
interested parties can send me mail. I think an architecture discussion is
probably the best place to start. In terms of cost, I want to make sure that
the PC board is available for less than $100.00 (not including nasties like
Tax, etc). A rough guess, is we probably need to make about 50 of the things
to be able to do them for that price. This, again, is an 'educated' guess.
Anyone with better figures should send me mail - figure that the board is,
at a minimum, four layers. I have access to a Mentor Graphics system, which
could be used for doing the PC layout, etc. Note, however, that a prototype
has to be built, which should be included in the cost. Comments?
- Der
--
dtynan@sultra.UUCP (Dermot Tynan @ Tynan Computers)
{mips,pyramid}!sultra!dtynan
--- God invented alcohol to keep the Irish from taking over the planet ---c60a-1cu@web-1e.berkeley.edu (Drew Dean) (11/10/88)
Ok guys, I'm one of those poor starving students -- I bought a Mac SE (through a Certified Developer, ie. discounted) last October over an AT&T Unix PC because I realized something: As much as I like to "hack" (better use quotes these days :-)), as a student, I spend a LOT of time doing other things, like word processing. MS Word on the Mac is much better for getting a paper out the door than the standard Unix tools. (My opinion, but near WYSIWYG vs dot commands means I only print things once, instead of finishing a paper at 10pm and not gettting any sleep because I can't get the print right...) I do a lot of modeming, and there's good software for that, too. Yes, I would like a 32 bit Unix box, (typing this from one of UCB's Sun 3/50's), but it better be under $1500....This eliminates the '532, as I just don't think it can be done....Now as to what I'd do: 1) Use a cheaper processor (ie 32332, but I don't know what a 16Mhz part costs), running at about 16Mhz. Remember, this is a single user system, and most of the time is wasted spent on the computer waiting for me to type something. 2) Put 2Mb RAM (of 256K chips) on the motherboard, and provide for expansion to 8Mb via 1M chips. 3) Put a SCSI port on the thing, as SCSI drives are getting cheap. (Volume from Mac sales may have something to do with this, all I know is that at the retail level, 20-40Mb SCSI drives have fallen $200-400 (roughly proportional to capacity), in the last year. 2 mail order places I know of are selling Quantum's 40Mb, 19ms drive with mouting hardware and Mac software for $650. Also, make sure the system will work well with a 40Mb drive.) 4) Leave the user the option of using an existing computer/terminal as the system console, or using a PC clone keyboard and a VGA monochrome or NEC Multisync GS monitor (ie. 640 * 480 or 800 * 600 monochrome graphics on a 14" screen). 5) Get a solid Unix on this system...If GNU arrives in reasonable working condition, that would be great...However, GCC still has some bugs (read gnu.gcc.bugs), and I think it will take another 6 months to 1 year for GCC to stablize, and about the same time for the GNU kernel (even if it's Mach) to become really usable. If GNU isn't there, National should step in by offering cheap binary licenses (4.3 BSD would be WONDERFUL, but I'll take Sys V). National -- tell the MBA's that you'll be getting a whole bunch of EECS students across the nation familiar with NS products, and look how well Apple's done in the last couple of years :-) Even in this group, I think it's safe to say that the NS32xxx series has not been a big commercial success for National Semi, although it does have technical merit. Conventional marketing techniques haven't worked very well, so why don't you try something a little different.... Drew Dean Internet: c60a-1cu@web.berkeley.edu UUCP: ...!ucbvax!web!c60a-1cu FROM Disclaimers IMPORT StandardDisclaimer;
stevel@eleazar.dartmouth.edu (Steve Ligett) (11/10/88)
I'd enjoy working on a project to build a 32k system, but I'd rather not start out by spending $532 for the chip set. Is anyone interested in starting with something cheaper? Here's all the pricing info I have: 532DK '532 designer kit $532 NS32332U-15 '332 cpu $177.45 NS32032E-10 '032 cpu $ 88.70 NSV-32016SK5-6 '016 designer kit $ 59.00 NSV-32032SK5-6 '032 designer kit $ 75.00 The two cpu prices are single piece prices from a salesperson at Time Electronics (an NSC distrib). Neither he nor I are familiar with the chips, so I didn't spend a lot of time trying to price a complete chip set, I was just trying to find a ballpark. The two designer kits prices are from a piece of literature that I got at Electro/88 in Boston at the national Booth (or the booth of an NSC distrib). The Time guy "didn't have them loaded into his computer". So, is anything less than a '532 interesting? The '332 chip is a little more expensive than a 16MHz 68020 but looks like it might be more affordable than a '532. Is there a '332 designer kit? I recall discussing a computer configuration some time ago under the subject of "532 Manifesto". Are we restarting that discussion? Steve Ligett steve.ligett@dartmouth.edu or Expiring minds (decvax harvard linus true)!dartvax!steve.ligett want to know.
dsb@Rational.COM (David S. Bakin) (11/10/88)
I missed the first part of this discussion, including the announcement (I suppose) of the 532DK. Can someone mail me the announcement if it in fact was sent on this group? I'd also like to know how fast the '532 chip that comes with the 532DK is. Thanks! -- Dave ---------------------------------------------------------- Dave Bakin (408) 496-3600 c/o Rational; 3320 Scott Blvd.; Santa Clara, CA 95054-3197 Internet: dsb@rational.com Uucp: ...!uunet!igor!dsb
dlr@daver.UUCP (Dave Rand) (11/10/88)
In article <2629@sultra.UUCP> dtynan@sultra.UUCP (Der Tynan) writes: > >Basically, it would be >a 'blank' printed-ciruit board for the IBM-PC. No components, nothing. >The design would allow for ALL 5 32032 chips (TCU, ICU, FPU, MMU & CPU). >Preferably running at something like 10MHz, but slower is OK, too. The >board would have space for up 8Mb of SIMM's (although it would run with less). > [text deleted] Now you're talking. How about: A 10 Mhz 32016 cluster (as above), 2 megabytes of RAM (perhaps hackable to 8 Mb), and 2 PROMS. Full schematics and PCB artwork available now. System V release 2 up and running. This was written up in Micro Cornucopia about 2 years ago. If someone wants to do a run of PCB's, I can help with the artwork... Dave -- Dave Rand {pyramid|hoptoad|sun|vsi1}!daver!dlr
josef@ugun21.UUCP (11/10/88)
In a previous note, George B. Smith (gbs@stratus.UUCP) writes: > As for an assembler, there is a "simple" > public domain 32k assembler written in C which runs on DOS available > from Rick Rodman of the NS32 User Group. As I am very interested in the NS32k, how could I get this assembler? Josef Moellers paper mail: e-mail: c/o Nixdorf Computer AG USA: uunet!linus!nixbur!nixpbe!mollers.pad Abt. EG-3 !USA: mcvax!unido!nixpbe!mollers.pad Unterer Frankfurter Weg D-4790 Paderborn tel.: (+49) 5251 104691 Standard disclaimer: Blablabla opinion blablabla employer blablabla!
adh@anumb.UUCP (a.d.hay) (11/10/88)
In article <7643@nsc.nsc.com> grenley@nsc.nsc.com.UUCP (George Grenley) writes: >In article <1264@cfa.cfa.harvard.EDU> ward@cfa.harvard.EDU (Steve Ward) writes: -->>There does seem to be active interest in a low-cost approach to -->>implementing a 32K machine of some sort. The idea of making it -->>compatible with PC hardware is a great way to keep the cost down. -->>There is all that chassis and peripheral and interface hardware -->>compatible with the PC/XT bus to be used, and many people probably -->>have such hardware. -->>It seems to me that the only to get such stuff is to make it ourselves. -->Hallelujah! We at NSC applaud you. Please keep me/us posted, we will -->do what we can to help. You can email to me at nsc!grenley does this mean you'll give Midnight Engineering Services a home? (;^>) -- Andrew Hay +------------------------------------------------------+ Holistic Specialist | I will design a computer for you, so powerful that | AT&T-BL Ward Hill MA | organic life will form part of its operational matrix| mvuxq.att.com!adh +------------------------------------------------------+
sedwards@esunix.UUCP (Scott Edwards) (11/11/88)
In article <2621@sultra.UUCP> dtynan@sultra.UUCP (Der Tynan) writes: >Another (hairbrained) idea would be to produce a '332 (or even '532) mother- >board for the IBM-PC. I mean, you can pick up individual clone parts (PSU, >case etc) real cheap these days. So, you have a motherboard which has up to >8Mb of SIMMS, and a lookalike PC-AT bus, so you can put off the shelf AT-type >plug-ins into the box. A friend of mine and I worked on building something like this last spring (with a 32016 because of $$$), but put it on the back burner because of a perceived lack of interest. I am willing to discuss taking it up again if anyone is interested, please e-mail me. I have an old 32016 system (6-Mhz) that is plenty fast for most things I have done at home. Let me know -- Scott
u5565522@ucsvc.unimelb.edu.au (David Clunie) (11/11/88)
It is with much interest that I have read all these passionate neew postings
directed towards producing a high-powered affordable board/system based on
the 32k family.
I too have wanted to do this for years. So much so that I even scrounged
one of those Compupro 32016 boards for my old S100 (non-Compupro) system.
It was easy enough to re-PAL this board to run with the TDS PROMS and put
in the extra chips (ICU,FPU) from the 32016 designers kit - and there you
have it a working board for about $A200 if you already have an S100 system.
But ... and this is a very big but ... who wants to then write there own
assembler, compiler, and operating system !!
Well, I started with a cross-assembler - simple two pass design with yacc
doing all the hard work to try to emulate the "standard" NatSemi syntax, and
got it to the stage of producing the correct byte codes, even in a
relocatable form. Got bogged down on producing COFF format output files
because the documentation in the NS manuals (like all UNix SysV references
to COFF) are a bit vague. Never got around to implementing branch
optimization, though I had an algorithm lined up for it.
The C compiler I am still working on, trying to keep up with revisions in
the ANSI draft standard, and developments in code generator technology.
This is one part of the project I am determined to finish, as it is
intended to be easily retargeted, and hence is generally useful even if
I loose interest in the NS family.
As far as an operating system is concerned there are two real alternatives:
- Minix - I have previously posted the results of my enquiries into
who has worked on Minix ports to 32k systems, and will repost them
if anyone is interested - the bottom line is that there are working
ports around.
However - Minix has major limitations - despite a very large amount
of software of the tool variety available from Minix users, you
won't be able to get any major applications (WP,database, etc)
available in source or binary form to run under MINIX as it is
presently conceived. To my mind, to achieve this on wants binary
compatibility with an existing commercial system ie. SysV from NS.
- Unix System V - is available at I understand reltively low cost
in binary form for various systems. Needless to say none of us are
going to be able to afford a source licence to write a port to our
own configuration from scratch. Hence any low cost board we design
and build is going to have to very accurately emulate something
that already exists ... point taken ?
Where have I got to ... basically, like many others here I have been
pottering around with the same idea for years ... trying to get a high
powerered 32k family machine and software working at rock-bottom cost.
Why have I persisted ? Well I guess the 32k architecture has this sort of
aesthetic appeal that the Intel and even Motorola families don't. Where
has it got me ? I now far more about writing compilers than I ever knew
possible and am thinking of making a living out of it. Just as well I have
another job ! I still can't sit down at my S100 system and run Unix on a
32016 so I guess I have wasted a lot of time. I have just bought a fast
AT clone to work on my C compiler faster, so I guess I have in some
ways succumbed to the segmented mob.
However I have not given up. If I can be of any assistance to anyone
developing an add in board for the PC using the 32k family that can be
assembled more cheaply than a Definicon or some such board can be bought
then well and good. I really think this is the way to go. Why ? Because
everyone needs a working development system NOW, not when they have finished
designing a new board. In other words everyone has a PC or a Mac. It is
too difficult to add to the Mac, and it costs to much for everyone to go
out and by another box,psu,hard drive,terminal etc. There is no other
way, it would seem to me.
Is it feasible ? There are several commercial products avaiable in a PC
add-in form. These use lots of gate arrays and have the facility to add
lots of on board 32bit bus memory. They run Unix with the host PC playing
IO processor. Are they any good ? I don't know I can't afford one. Can we
produce a cheaper, do it yourself kit ? Maybe - but I would suggest a careful
analysis of wherein the cost of these boards lie - it may be that there is
just so much glue hardware, plus the chip cost, plus fabrication of a
multilayer PC board, plus memory in large quantities - that we may not
be able to do it any cheaper.
Well - enough. I don't want to seem pessimistic, but one has to face reality.
Happy to be of service to anyone who needs it ...
Regards ... Davidmlewis@unocss.UUCP (Marcus S. Lewis) (11/12/88)
In article <7259@daver.UUCP>, dlr@daver.UUCP (Dave Rand) writes: > Now you're talking. > How about: A 10 Mhz 32016 cluster (as above), 2 megabytes of RAM > (perhaps hackable to 8 Mb), and 2 PROMS. Full schematics and PCB > artwork available now. System V release 2 up and running. > This was written up in Micro Cornucopia about 2 years ago. If someone > wants to do a run of PCB's, I can help with the artwork... The question is, why didn't the PD-32 ever get made? I understood Definicon took it over and was willing to distribute the boards, but they had serious problems with the boards/prototypes. Or was that just problems with marketing a competing product? I don't want Unix. I want a NSC32. The last word I hrd from Definicon was they would sell the PD-32 real cheap, but to get the cheap price you had to buy Unix. I just want hardware. I also don't have a "PC", although for the right add-on, I might swing it. Marc
boing@cidam.rmit.oz (Wild Bill Vurtoch) (11/12/88)
> > Gee, the cg16 group is not going to like you calling there new > baby hairbrained. :-) > Can anybody tell me what a CG16 is ??? What do you/will you get with this product.
george@wombat.UUCP (George Scolaro) (11/14/88)
In article <2944@cidam.rmit.oz> boing@cidam.rmit.oz (Wild Bill Vurtoch) writes: >> >> Gee, the cg16 group is not going to like you calling there new >> baby hairbrained. :-) >> >Can anybody tell me what a CG16 is ??? What do you/will you get with >this product. ^^^^^^^^^^^^^^^^^^^^^^^^^ Will you get? buy the chip, its available now. Do you get?... read on. The CG16 or NS32CG16V-10, -15 is a 68 pin plcc device that contains the 32C201 (TCU) and 32C016 (CPU) functionality on one chip. It is a CMOS device. The instruction set is 100% (+/-) compatible with the rest of the 32000 series of CPUs. It does NOT have the MMU instructions, eg MOVUS etc and does NOT connect to an MMU. It does HAVE extra instructions to assist in its intended application, ie graphics. It has BITBLT instructions and some RLL primitives. Think of the NS32CG16 as an embedded controller, ie its cheap, around US$20 - 30 and requires the addition of a crystal, or module... It DOES interface to the 32081 or 32381 FPU, with the 32381 it DOES support the new transcendental primitives. I have built a simple CG16 board that runs at 15 Mhz, uses 4 256k x 4 bit 120 ns drams, a NS32CG821 (new version of DP8421), DUART, couple of gates, buffers and latches and 2 256k bit eproms, and an e2rom. The total chip count is 27 devices. It interfaces to one of National's demo ethernet boards. It is basically a printer buffer, ie it takes either of the serial uart channels, and or the ethernet channel, gives 512k bytes of buffer and talks via it's the centronics port, to a laserjet II. Its getting cold and miserable (silcon valley) now, so one has to do something in the winter to keep busy. G'day from W.A. (via the U.S.A.)