STEINER@RUTGERS.ARPA (01/07/85)
From: Caro.PA@XEROX.ARPA Does anyone manufacture a UNIX box based on the NS320xx?? If so, what are the specs? I like the chip, but everyone seems to have jumped on the 68K bandwagon. I read an add that Tektronix (I think) is coming out with a CAD/CAM workstation based on a 32032/32016 coprocessor design, but I don't think it's UNIX based. Any info is welcome. Perry
STEINER@RUTGERS.ARPA (01/07/85)
From: Doug <Faunt%hplabs.csnet@csnet-relay.arpa> American Information Systems (415)494-3210 is shipping a 32032 based Unix system, any day now, so I'm told. faunt%hplabs@csnet-relay ....!hplabs!faunt
iau@ukc.UUCP (I.A.Utting) (01/08/85)
An outfit called "Whitechapel Computer Works" in London have a workstation (called the MG-1) based on the NS 32016 running GENIX (Nat Semis 4.1 port). It has a monochrome bit-map display (1024 x 800), 40MByte Winchester, 5.25" floppy, a Swiss 3-button mouse, on-board Ethernet and 1MByte RAM. It also has an IBM-PC expansion bus (DMA capable) which should keep the price of peripherals down. All this for (pounds sterling) 8000 or there abouts. I have a beta-test machine which is behaving reasonably for such things, but it's early days yet for software which takes advantage of the screen (window manager soon) or the Ethernet (no support yet). You can contact WCW at: Whitechapel Computer Works 75, Whitechapel Road London E1 1DU UK Phone: +1 377 8680 Ian.
ed@mtxinu.UUCP (Ed Gould) (01/09/85)
> From: Caro.PA@XEROX.ARPA > > Does anyone manufacture a UNIX box based on the NS320xx?? If so, > what are the specs? Sequent Computer Systems in Portland recently announced a multiprocessor system, containing from two to 12 32000 CPUs. It runs 4.2BSD. I understand that they'll be showing it in Dallas. It's a nice machine. -- Ed Gould mt Xinu, 739 Allston Way, Berkeley, CA 94710 USA {ucbvax,decvax}!mtxinu!ed +1 415 644 0146 (I'd rather not be parochial.)
STEINER@RUTGERS.ARPA (01/09/85)
From: Andrew Klossner <andrew@orca> "Does anyone manufacture a UNIX box based on the NS320xx?? If so, what are the specs? I like the chip, but everyone seems to have jumped on the 68K bandwagon. I read an add that Tektronix (I think) is coming out with a CAD/CAM workstation based on a 32032/32016 coprocessor design, but I don't think it's UNIX based." Our "CAD/CAM workstation" is just a 320xx-based computer running a 4.2BSD port. I have one on my desk; it's a very nice little box. We added mucho bells and whistles, like a distributed file system (my neighbor's disk looks like part of my file system), the "unimplemented" memory management calls like mmap(2) and munmap(2), and, since we're Tektronix, a real Basic language implementation. -- Andrew Klossner (decvax!tektronix!orca!andrew) [UUCP] (orca!andrew.tektronix@csnet-relay) [ARPA]
jps@stcvax.UUCP (Jeff Snover) (01/10/85)
> Does anyone manufacture a UNIX box based on the NS320xx?? If so, > what are the specs? Intergraph (Huntsville Alabama) has a nice workstation based on the 32032 called the Interpro-32. Here is what I remember: - 32032 based running Genix (nice Unix [4.2?] port) - 32032 support chips (I think the MMU and FPP) - 1k x 1k color display with windowing - 1.75 Megs expandable to 4 (or 16) Megs - Intel Ethernet controller chip coupled with the 80186 (yuk) to handle packets and I/O. - *fast* bit-slice engine to handle screen manipulation and graphics - 25 Meg Winchester I don't know if this is appropriate for your application but I was quite impressed. I hope this was useful. (Oh by the way, my brother is a Civil Eng. and a buddy of his works at Stone & Webster and said of the *many* workstations that they have [Apollo, Computervision, Camdus(???), and Intergraph] he has found the Intergraph to be the best). ------- -- Chapter-11/Multiple-Mega-Layoff Survivor (for now...) Jeffrey P. Snover - STC StorageTek (Disk Division) uucp: { ihnp4, decvax}!stcvax!jps { allegra, amd70, ucbvax }!nbires!stcvax!jps USnail: Storage Technology Corp - MD 3T / Louisville, CO / 80028 DDD: (303) 673-6750
STEINER@RUTGERS.ARPA (01/15/85)
From: psu-cs!aatpdx!mcg%tektronix.csnet@csnet-relay.arpa As a former employee of Tek working on the workstation to which you referred in your note, I can tell you that it is *definitely* UNIX based. The system (the 6000 series) is based on the 32016 and 32032 processors and 4.2BSD UNIX. I don't wish to say anything about their merit. Also, a company called LMC Corp sells a Multibus-based 32016 box which runs UNIX. There are two commercially available UNIX ports for the 32016: one from Human Computing Resources in Toronto: originally a 4.1 port, now moving to System V; and one from National Semiconductor, also originally 4.1, now moving to 4.2. HCR's is available in source or binary form for a small number of configurations, and National's is available in source form, or in binary for their proprietary workstation. I run HCR 4.1 UNIX (Unity) on both an old National DB16000 board (a multibus 16032 prototype) and on the GVC Corp GVC-16 board, made by a small company in Cambridge. Since we do not yet have Rev N (bug-free) chips, all the software work-arounds make it a little slow, but I expect 11/750 performance or better when we have Rev. N chips and use GVC's 4 megs of on-board no-wait-state memory (there are 4 to 6 waits for Multibus memory access). S. McGeady Ann Arbor Terminals Research and Development Portland, OR
shor@sphinx.UChicago.UUCP (Melinda Shore) (01/17/85)
Sad rumor time -- I hear LMC has gone under, or at least has been dropped by their parent company. The Unix they've been running, is Genix, which is based on 4.1. It really looked like a nifty machine. I hope that they can find other sources of capital. Melinda Shore University of Chicago Computation Center
STEINER@RUTGERS.ARPA (01/17/85)
From: Tom Blenko <blenko@rochester.arpa> The Tektronix box does run on a 16/32K. But the 16K only runs at 4 MHz (which they all somehow forgot to mention in the WORKS digest). Look before you leap! Tom
steveg@hammer.UUCP (Steve Glaser) (01/19/85)
>From: Tom Blenko <blenko@rochester.arpa> > >The Tektronix box does run on a 16/32K. But the 16K only runs at >4 MHz (which they all somehow forgot to mention in the WORKS digest). >Look before you leap! > > Tom Now wait just a minute here. Currently Tek 6000 workstations run at 8Mhz. We will be moving to 10Mhz as soon as the chips are available in sufficient quantity. The 10 Mhz systems we have in house work just fine. We never did run at 4 Mhz. There were some early software development and demo boxes that ran at 5 Mhz. Steve Glaser Tektronix, ECS Engineering tektronix!steveg UUCP steveg.tektronix@csnet-relay CSNET/ARPANET
steveh@hammer.UUCP (Stephen Hemminger) (01/20/85)
----- | The Tektronix box does run on a 16/32K. But the 16K only runs at | 4 MHz (which they all somehow forgot to mention in the WORKS digest). | Look before you leap! ----- Wrong! Production TEK workstations run at 8 Mhz. Your information may have come from software vendors who were loaned pre-production test units at lower clock rates.
STEINER@RUTGERS.ARPA (01/21/85)
From: Andrew Klossner <andrew@orca>
Tom Blenko <blenko@rochester.arpa> writes:
"The Tektronix box does run on a 16/32K. But the 16K only runs
at 4 MHz (which they all somehow forgot to mention in the WORKS
digest). Look before you leap!"
The Tektronix 6000 family of workstations (NS16032/32032 based, running
enhanced 4.2BSD Unix) presently run at 8MHz, and will increase to 10MHz
in the next month or so as the rev N 32016 CPUs become available.
These products have never run at 4MHz, even during initial development.
The first engineering prototypes clocked at 5MHz, back when National
couldn't produce parts faster than this.
As Mr. Blenko is a former employee of Tektronix, I'm surprised at his
vociferousness.
-- Andrew Klossner (decvax!tektronix!orca!andrew) [UUCP]
(orca!andrew.tektronix@csnet-relay) [ARPA]
STEINER@RUTGERS.ARPA (01/22/85)
From: psu-cs!aatpdx!mcg%tektronix.csnet@csnet-relay.arpa Tom Blenko (blenko@rochester) is completely wrong in the statement that "the Tektronix box .. only runs at 4 MHz". Early prototype versions of the box ran with 4 and 6Mhz parts, because of the inavailability of 10Mhz ones, and bug-free 10Mhz parts (rev N) are still not available in production quantities, but even though I haven't worked for Tek for over a year, I know that the first production units will run at least 6 Mhz, and later units will run at 10 Mhz. Tom is also a former Tek employee and obviously has some axe to grind. Since the product has been announced and is almost on the market, and has been the subject of a certain amount of bashing, let me reverse my previous stand of silence on the issue, and say this: the Tek 6000 series will be a very nice UNIX box. Engineering Computing Systems (ECS, the division building the system) has expended an amazing amount of time and effort fixing bugs in the system, rationalizing interfaces, and generally compiling all the best features of numerous UNIX systems out there for their version. They have made some major strides forward in improving the virtual memory system in 4.2bsd, and have sped up the system by off-loading major subsystems into other processors - the I/O system and network are supported by independent processors. Take the Tektronix box seriously. S. McGeady Ann Arbor Terminals Portland, OR cbosgd!aat!mcg
tucker@ccvaxa.UUCP (01/27/85)
Whitechapel in the UK makes a real nice 32K UNIX single user workstation from the sounds of it (See the Fe issue of BYTE magazine). Claims to use some very advanced hardware ideas like: 1.) A wide data bus (64bits) with 200Mhz bandwidth (bit bandwidth). 512K standard, 8Meg max. 2.) Two coprocessors for the bit mapped display, one for BLT's and one the keyboard/mouse and mouse icon. I.e. main processor time is used only when something happens at the user interface. Just running the mouse around doesn't take up any main processor time. Also, the display processor uses virtual memory mapping with its own page table handler. I.e. display memory can be non-contig and and as many displays as the uses wants can be used at once. 3.) Hard-disk has a direct connection to the memory that doesn't take time away from the main CPU because of high memory bandwidth. Hard-disk DMA allows multisector unloading to noncontig locations without CPU time being wasted. 10/22/45 Meg, also one 800K floppy. Other nice hardware things: a.) Ethernet interface standard. b.) Full 8Mhz 32000 processor set standard. c.) Nice landscape display with keyboard and hemisphere mouse standard. d.) Soft on/off switch like that on the Apple LISA standard. They also have some nice software things: 1.) BSD 4.1 Genix optimized for workstation use. (i.e single user) 2.) Comes GKS software and windowing built in. 3.) Full C, PASCAL, and FORTRAN optimizing compilers available. (??) 4.) Smalltalk maybe (vaperware) available in the near future. The nicest thing of all was the price. 5000lbs for the base configuration. Thats about $6000 dallors at current exchange rates I think. Sounds too good to be true. Note that I'm not affiliated with this company, and do not know details about the system. Look at BYTE if you want to see the reveiw.
jml@drutx.UUCP (LeonJM) (01/30/85)
National has recently announced a System V Rel 2 version 2 port. It is the first System V port on a micro chip that supports paging. I played around with a system at the National booth at UniForum and it seemed reasonably fast. By the way, does anyone know the who actually is doing the official System V port for Zilog? John Leon ihnp4!drutx!jml AT&T Information Systems Laboratories, Denver
sohail@terak.UUCP (Sohail M. Hussain) (02/01/85)
> Early prototype versions of > the box ran with 4 and 6Mhz parts, because of the inavailability of 10Mhz > ones, and bug-free 10Mhz parts (rev N) are still not available in production > quantities We have some of those 10Mhz rev N parts, in our work station and what has been puzzeling me, is that these machines out perform out Vax 750. (not in compiles ofcourse, but in execution times) Can some one out there shed some light on why a 32016, runs faster than a 750, in programs that access memory (using pointers or matrix type operations.) As background, we have Vax 750, with 3Mb mem, running 4.1 BSD and the workstations are of our own manufacture, using a 10Mhz 32016, 4Mb memory, and running 4.1 Genix. The times were done with both systems running multiuser, but only one person logged in. Looking forward to some answer, as I have always though of our 750 as a fair size machine, it supports our development efforts quite well, I am now faced with either having to start respecting the 32016 more, or the 750 less. sohail Sohail Hussain uucp: ...{decvax,hao,ihnp4,seismo}!noao!terak!sohail phone: 602 998 4800 us mail: Terak Corporation, 14151 N 76th street, Scottsdale, AZ 85260 -- Sohail Hussain uucp: ...{decvax,hao,ihnp4,seismo}!noao!terak!sohail phone: 602 998 4800 us mail: Terak Corporation, 14151 N 76th street, Scottsdale, AZ 85260
hammond@petrus.UUCP (02/07/85)
> We have some of those 10Mhz rev N parts, in our work station and > what has been puzzeling me, is that these machines out perform out > Vax 750. (not in compiles ofcourse, but in execution times) > > Can some one out there shed some light on why a 32016, runs faster > than a 750, in programs that access memory (using pointers or matrix type > operations.) > > Sohail Hussain > Issues: Does your 32016 based workstation have a 32081? Are you using the 32082 MMU? Does your 750 have a floating point accelerator? Is your benchmark program small enough to fit in memory, (i.e. roughly the same number of page faults on both machines?) Questions: How much faster, i.e. 5, 10, 20 30 %? I have a NSC Sys 32 (A 32016 based, 4.1 bsd development system) It runs about the same as an 11/23, or about 1/3 of a 750. My boss has been giving me grief about this, so your info is most encouraging. Note a 32032 should give roughly 1.25 times the performance of a 32016. The 32 bit bus doesn't buy you that much more, except in applications such as copying data memory to memory.
henry@utzoo.UUCP (Henry Spencer) (02/08/85)
Another relevant question is, does your memory have zero wait states? People I trust tell me that the 32016's performance deteriorates *SHARPLY* when wait states are introduced -- it's much worse than you would expect, and in particular it's not linear in the number of wait states. -- Henry Spencer @ U of Toronto Zoology {allegra,ihnp4,linus,decvax}!utzoo!henry
chuqui@nsc.UUCP (Chuq Von Rospach) (02/09/85)
In article <278@petrus.UUCP> hammond@petrus.UUCP writes: >It runs about the same as an 11/23, or about 1/3 of a 750. >My boss has been giving me grief about this, so your info is most >encouraging. I'll probably get grief for saying this, but there are some quirks in the SYS32 hardware that keep it from performing in ways it should. The memory subsystem tends to require an unreasonable number of wait states in certain configurations, and it makes the system sludge out. We've been taking a close look at the SYS32 in the last few months because we realize that the performance makes our chips look a lot worse than they really are. I don't have anything I can talk about at this time besides pointing out that it IS very possible to get 32xxx based systems that run MUCH faster than SYS32. The SYS32 is more of a workhorse than a benchmark system, and people should be aware of that fact. chuq -- From the ministry of silly talks: Chuq Von Rospach {allegra,cbosgd,hplabs,ihnp4,seismo}!nsc!chuqui nsc!chuqui@decwrl.ARPA Life, the Universe, and lots of other stuff is a trademark of AT&T Bell Labs
srm@nsc.UUCP (Richard Mateosian) (02/11/85)
In article <5040@utzoo.UUCP> henry@utzoo.UUCP (Henry Spencer) writes: >Another relevant question is, does your memory have zero wait states? >People I trust tell me that the 32016's performance deteriorates >*SHARPLY* when wait states are introduced -- it's much worse than >you would expect, and in particular it's not linear in the number of >wait states. Obviously, behavior under wait states depends a lot on particular programs. Here is one data point taken from one of my Wescon papers. The program is a small benchmark that uses memory heavily. Bus use is 82% on the NS32016, 57% on the NS32032, both at 0 wait states. Given below are execution times in seconds at 0, 1 and 2 wait states: 0 ws 1 ws 2 ws ---- ---- ---- NS32032 32.7 36.0 39.3 NS32016 43.7 49.4 55.0 Ratio 1.34 1.37 1.4 Execution speed of the NS32016 is 88.5% at 1 ws, 79.5% at 2 ws. Execution speed of the NS32032 is 90.8% at 1 ws, 83.2% at 2 ws. In general, programs with lighter bus use show smaller degradation with wait states and smaller ratios of NS32032 to NS32016 execution speed. In fact, a program doing nothing but register to register divison operations might show no degradation at all under wait states (because of the instruction pipeline) and no difference in execution speed between the NS32016 and NS32032. -- Richard Mateosian {allegra,cbosgd,decwrl,hplabs,ihnp4,seismo}!nsc!srm nsc!srm@decwrl.ARPA
jchapman@watcgl.UUCP (john chapman) (02/12/85)
> Another relevant question is, does your memory have zero wait states? > People I trust tell me that the 32016's performance deteriorates > *SHARPLY* when wait states are introduced -- it's much worse than > you would expect, and in particular it's not linear in the number of > wait states. > On the otherhand (at least according to my old 16032 manual) you wouldn't need very fast memory to keep up with a 32016 say about 400ns access?
doug@terak.UUCP (Doug Pardee) (02/12/85)
> >People I trust tell me that the 32016's performance deteriorates > >*SHARPLY* when wait states are introduced -- it's much worse than > >you would expect, and in particular it's not linear in the number of > >wait states. > > In general, programs with lighter bus use show smaller degradation with wait > states and smaller ratios of NS32032 to NS32016 execution speed. Wait states are a punch aimed at the 32000's glass jaw -- instruction prefetch. For those not completely conversant: the 32000 series CPU's use instruction prefetching to try to keep the 8 bytes following the _current_ instruction already loaded into the CPU. These bytes are always the ones located sequentially after the current instruction. There are two undesirable side effects which can occur. The most obvious occurs when a branch is taken -- the prefetch cycles were a waste of time, and the new instructions have to be fetched. But ---- if the CPU had just started a prefetch cycle when the branch is recognized, it has to wait for it to complete before the branch can be executed. Wait states increase the likelihood of this happening as well as make the situation more serious. Remembering that programs spend most of their time in loops, and that a loop requires at least one branch on every time through, this effect is magnified considerably. Especially for concocted benchmark programs, where the contents of the loop tends to be trivial, leaving the branching as the major time consumer. A second aspect of the 32000 series enters in here as well -- unlike the 68000, instructions are not required to start on word boundaries. If the branch destination is to an "odd" address, the CPU requires yet another memory cycle, with any wait states. Compilers for high- level languages like "C" don't pay any attention to this little detail, so tight loops can suffer just because the top of the loop is on an odd-byte boundary. The other side effect is less obvious. The instruction prefetch cycle can also obstruct access to the operands of the current instruction. Again, wait states increase the likelihood of this happening, and make the delay more serious as well. This process, in turn, is made more likely by the use of high-level languages like "C". Unlike the competition's CPUs, the 32000 series allows essentially all operations to be performed memory-to-memory, without needing a register as an intermediate. The compilers use this feature extensively, with the result that operands require memory access much more often than the equivalent 32000 assembler code or (e.g.) 68000 "C" code. Important note: this presumes that if the compiler had been forced to bring the operands into a register, and get the result in a register, that it could have done some optimization and re-used that register. It is obvious, is it not, that a simple "Load A, Add B, Store B" is necessarily going to be slower than "Add A to B"? And to compound the problem even further: the 32000 series is set up to use "indirect addressing" fairly heavily, and the compilers really use it a bunch. Especially the "C" compiler, which uses indirect addressing to implement pointer variables. But wait, there's more (this is starting to sound like a TV mail-order ad!). Most "C" programmers seem to like to use "external" variables rather than parameters. On the 32000 series, parameters are accessed just as easily as ordinary variables, but externals are a *double- indirect*! For a 32016 to get just the *address* of an external item, it has to do four (4) memory cycles. And if that item is a pointer variable, "C" will require yet another two memory cycles before it even has the *address* of the data. All of this indirect address and operand fetching puts quite a load on the memory system, and prefetching represents serious competition for memory cycles. If that prefetching turns out to have been unnecessary because of a branch, the performance suffers more than the number of wait states would imply. So if you want your 32000 system to hum along, don't use wait states, keep looping and branching to a minimum, program in assembler, and if you simply *must* program in "C" avoid external variables and use register variables (especially for pointer variables). Oh, BTW, the MMU adds one wait state of its own. -- Doug Pardee -- Terak Corp. -- !{hao,ihnp4,decvax}!noao!terak!doug