scwilk@sdrc.UUCP (Ken_Wilkinson) (11/24/88)
Dear net folks,
I have watched with more than just some amusement at the discussion
that has followed the reposting of my '532 Manifesto. I think its time
that I reiterate the "VISION" that I had with the original posting.
<FLAME ON>
BUT FIRST! lets cut out the crap on the OS discussion please!
I'm sure that you can discuss the merits of your favorite OS
in a appropriate group. We don't (in case you haven't noticed)
have ANY license. If you want MINIX do it! NOW! I personallly
would like a new toy to play with but calling Minix a real os
is like calling Pascal a real programming language. :-)
(flame suit is on)
<FLAME OFF>
On another note, I would like to make a list of anyone or everyone who
is just more than mildly interested in this discussion. So please fill
out the following questionaire, and I'll post the results.
Question #1
The discussion on the '532 Manifesto interest me as far as
participation is concerned...
a) very interested, willing to help in any way (Hwr/Sof)
b) somewhat interested, will help as much as I can time permitting
c) interesting to read messages
d) not interested.
e) get off the net turkey!
Question #2
Provide your name, address (net, land coordinates etc) and
phone number to be reached.
Question #3
Provide your expertise, hardware, software, 25 words or less resume.
Question #4
How much would you spend? (whats your threshold of pain?)
a) $1000
b) 1-$2000
c) 2-$3000
d) $50,000 (call me quick!)
e) other amount ____
Question #5
What would you use this beast for?
a) research
b) toy
c) replace your: wife/husband/girl/boy-friend/dog.
d) business
e) other ___________
Please E-mail your answers to the questions as:
#1, a; #2 b; etc.
This will keep the net trafic down.
The VISION!
The original idea was to build an inexpensive computer without competing
with the NeXT, Mac, IBM, SUN, etc. world. As such it was decided that
the computer could be based on a Single board, plugin board, or IBM pc
style motherboard. The purpose of using the IBM PC bus, motherboard
design was to allow the use of cheap periphials and take advantage of
someone's else volume of manufacturing. Thus the case, powersupply,
keyboard, display, etc could be used. What this means is we can get a
decent architecure running and have "reasonable" performance without
spending years and bucks to do it! You can tailor it to meet your needs.
If someone wants to go cheap, they could use a terminal (if they have one)
and the motherboard. Or, buy a hiz rez VGA 1024 x 782 16 bit card and
NEC multisync II screen. Still cheaper that a homebrew adapter I'm sure.
The PC bus is admittably not the best choice, personally I would like to
use the Nubus, especially with the new TI chips to interface to the Nubus.
BUT! what cheap periphials do you see for the MAC II? NONE, unless your
into the SuperMac display (see above) or IEEE-488 instrument control
(bet its cheaper to buy a pc board). Using the SCSI bus for the hard
drive on the motherboard would give decent hard disk performance which
is what is one of the most important things anyways. Memory however
is the other. I don't think that we need to use sub-nano second memory
to get decent performance. I we stick to mass produced and available
SIMMS which are getting used in the volume machines I think that we
could fit enough memory on the motherboard to satisfy all of us.
Probably 16 megs at least! YOU say you need more? Tuff, you got the
wrong computer buddy boy! You say you need less than 16 meg?
Fine, buy what you need...
I believe that if enough of you are willing, and a design can be agreed
upon, that we may see NSC step up to help. Lets try to firm this up
in the next week or so, and see who and how many respond to the
questionaire.
>> Just remember, as in any design, some trade-offs MUST occur. <<
---------------------------
P.S. How do you like the name? Sexy huh?
Kenneth Wilkinson
SDRC
Cincinnati, Ohio 45150
(513)-576-2569 {work}
(513)-831-9441 {home}
Net (uunet!sdrc!scwilk)
"Are we having fun yet?"
standard disclaimers applybga@raspail.UUCP (Bruce Albrecht) (11/28/88)
I'm not very hardware knowledgeable, so bear with me. I also don't have any '532 datasheets, although my local sales office has ordered the 1988 databook for me. If we were to use the 25 Mhz '532 out of the 532DK (can we do this? Can the '532 be removed from the 532DK PC board?), and 100 ns DRAM, are there any problems? I realize that any memory request would need wait states. Would we be able to use burst mode? The '532 has onboard cache. Wouldn't it have a large enough hit rate that the '532 would still have a very acceptable 2-3 MIPS rating? If we went this route, what sort of special hardware might we need?
curry@nsc.nsc.com (Ray Curry) (11/29/88)
In article <1061@raspail.UUCP> bga@raspail.UUCP (Bruce Albrecht) writes: >I'm not very hardware knowledgeable, so bear with me. I also don't have any >'532 datasheets, although my local sales office has ordered the 1988 databook >for me. > >If we were to use the 25 Mhz '532 out of the 532DK (can we do this? Can the >'532 be removed from the 532DK PC board?), and 100 ns DRAM, are there any >problems? I realize that any memory request would need wait states. Would >we be able to use burst mode? The '532 has onboard cache. Wouldn't it have >a large enough hit rate that the '532 would still have a very acceptable 2-3 >MIPS rating? If we went this route, what sort of special hardware might we >need? Perhaps I can answer several questions at the same time. The 532DK comes totally non-assembled including a socket for the NS32532 so the CPU is available. Secondly the '532 interfaces pretty well with 100 ns memory. If you don't have an external cache to have search a tag table for, the memory available time is two bclocks (80ns) less address valid (8ns) and less data input setup (11ns). This available time is distributed over address and data buffers as well as the memory access time. With each wait state, this increases by 40 ns. Depending upon the buffers used and decode technique, you should be able to run 100 ns memory with 2 waits. You can also consider running at 20MHz and get by with 1 wait, or use 80 ns for 1 wait at 25 MHz. To get the most performance either way, you would want to use burst memory access which uses one clock for each of the next 3 reads. To do this with your 100 ns memory, you would have to interleave memory banks so that while one bank is being read, the address is setup to the second bank. (Or consider using the nibble mode RAM from INMOS). As to performance under these circumstances, the main reason for the external cache on the VME532 board is to allow easy (slow) memory access to main memory. Running out of main memory on the VME532 board uses something like 7 waits. In running Dhrystones, I noticed about 10% degradation for the first wait state and am told by our Israeli chip experts that Dhrystones are a bad case, worse than average. My advice for a good but not absolutely highest performance UNIX or similar based machine would be to do just that, simplify the memory design, forget an external cache, and use burst mode. Another posting asked about whether the MMU took an extra 50 ns. Accesses with the MMU enabled do not take extra time as long as the translation table entry is in the MMU's TLB. Each time the CPU enters a new page table that doesn't not have an entry in the TLB, the MMU will do a look up in the external tables. The table entries stored in the MMU are the 64 most recently used. Since the pages are 4K, the effect of the table look ups are minimal.
curry@nsc.nsc.com (Ray Curry) (11/29/88)
Assuming this project is developing into something real, the committee should form a chairman, preferable someone connected with a college. Said chairman should then contact Linda Dimmick, in Series 32000 Applications/Marketing, since she runs our college program. Maybe we can work something out on some 32532B3 parts, less designer kit board.
curry@nsc.nsc.com (Ray Curry) (11/29/88)
Oops, I meant to include Linda Dimmick's phone number in my preivous posting, and didn't. Its 408-721-7295.