History of 1130 emulators
Eddy Quicksall
Here is a quiz before I give you the history. What do you think is the oldest 1130 emulator?
Eddy
Matt Patoray
Here is a quiz before I give you the history. What do you think is the oldest 1130 emulator?
Eddy
--
You received this message because you are subscribed to the Google Groups "IBM1130" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ibm1130+u...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
Owner, MSP Productions
KD8AMG
Eddy Quicksall
I don’t know about that one. I wonder what year that was.
Eddy Quicksall
From: ibm...@googlegroups.com [mailto:ibm...@googlegroups.com] On Behalf Of Matt Patoray
Sent: Friday, July 29, 2016 12:57 PM
To: ibm...@googlegroups.com
Subject: Re: [IBM1130] History of 1130 emulators
Isn't there a 1130 emulator in the System/3 OS?
Eddy Quicksall
Maybe. Do you have a reference?
From: ibm...@googlegroups.com [mailto:ibm...@googlegroups.com] On Behalf Of Matt Patoray
Sent: Friday, July 29, 2016 12:57 PM
To: ibm...@googlegroups.com
Subject: Re: [IBM1130] History of 1130 emulators
Isn't there a 1130 emulator in the System/3 OS?
John Doty
Here is a quiz before I give you the history. What do you think is the oldest 1130 emulator?
Eddy--
You received this message because you are subscribed to the Google Groups "IBM1130" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ibm1130+u...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
John Doty Noqsi Aerospace, Ltd.
Eddy Quicksall
I think what you are talking about is something that would test the instruction set. Actually I was talking about a full emulation that made one think he was running on an actual 1130.
Eddy
From: ibm...@googlegroups.com [mailto:ibm...@googlegroups.com] On Behalf Of John Doty
Sent: Friday, July 29, 2016 1:16 PM
To: ibm...@googlegroups.com
Subject: Re: [IBM1130] History of 1130 emulators
On Jul 29, 2016, at 12:13 PM, Eddy Quicksall <Ed...@Quicksall.com> wrote:
Eddy Quicksall
I forgot to tell you … this was a timed quiz and times up now.
Here is what I remember. Feel free to fill in your part if you wrote one too.
1971 – Digital Scientific built a computer intended to replace the 1130. It was called the DSC Meta-4. A guy named Dike Summers worked at DSC wrote the 1130 emulator to run on that machine. That was a cool time. The emulator is all on copper squares about 1/10th inch square (the bits). That is when the term firmware really meant something because it was certainly firm.
1972 – I was doing work for a guy in Miami (Richard J. Sherin). He observed that Dike’s emulator took 3 ROM modules and each module cost around $4000. So Richard wrote a replacement emulator that only took 2 ROM modules and Richard included “hardware” floating point with his. At DNA we used that emulator until the end of the 70’s. I have one of those 30 pound ROMs in my attic. I live in FL on the beaches and it is too hot to go up there. I’ll get one down sometime and show everyone a picture of it.
Mid 70’s – Another company tried to get in on the action. The CHI 2130 (really cool and looked like a small 360). It emulated the 1130 but I think he did all of that in hardware.
Mid 70’s – General Automation had built a computer to “replace” the 1130. It was the GA 18/30. But it was all hardware. It supported lots of extra memory and large disk shared drives. I didn’t know it at the time but that is what was the basis for my emulator. GA’s customers were getting mad because GA was not delivering a multi-tasking DM2 as the promised. So GA contracted DNA to put TSO and CYTOS on it.
Mid 70’s – I met a guy at the Common conferences that worked for IBM. He had written an 1130 emulator for some IBM computer. I don’t remember which one. But it was just something to play with and I don’t know if it was full featured.
1980 – The owner of DNA, Don Nichols, got a guy at Data General to build a module for the Series/1. The board gave extra power to support TSO (our bread and butter at DNA). So Don wrote yet another 1130 emulator for the Series/1. But no one cared since all the stack based computers took the limelight. So that one became a bust.
1994 – Around 1981 one of DNA’s customers bought all of the rights to the DNA systems and got all of the source. He had General Automation 18/30’s (a mix of the 1130 and 1800). He ran these 18/30’s until I showed up at his door. He told me he had a very high electric bill because he had to run his air conditioner in the winter and he lived in KS. So being a smart #$@% I told him I would replace his power hungry computers (they were clustered) with just a few PC’s. So he contracted me to write yet another 1130 emulator. He is still running those in everyday business supporting his customers with accounting, payroll, general ledger and other things I don’t understand.
Around 1996 – I met Brian and visited his office to see real 1130’s. He showed me an emulator he wrote but I don’t know when he wrote it.
Current – lots of other emulators. Please fill this in with your dates and short history.
From: ibm...@googlegroups.com [mailto:ibm...@googlegroups.com] On Behalf Of John Doty
Sent: Friday, July 29, 2016 1:16 PM
To: ibm...@googlegroups.com
Subject: Re: [IBM1130] History of 1130 emulators
On Jul 29, 2016, at 12:13 PM, Eddy Quicksall <Ed...@Quicksall.com> wrote:
John R Pierce
Mid 70’s – Another company tried to get in on the action. The CHI 2130 (really cool and looked like a small 360). It emulated the 1130 but I think he did all of that in hardware.
yeah, the CHI (Computer Hardware, Inc) 2130 CPU was all hard
wired TTL logic. Instruction cycle times were in the 800nS range
IIRC. I worked for CHI in 1975, and the 2130 was a done deal
and had been shipped to 100(?) or so customers.
-- john r pierce, recycling bits in santa cruz
Bob Flanders
Eddy Quicksall
I forgot to tell you … this was a timed quiz and times up now.
Here is what I remember. Feel free to fill in your part if you wrote one too.
1971 – Digital Scientific built a computer intended to replace the 1130. It was called the DSC Meta-4. A guy named Dike Summers worked at DSC wrote the 1130 emulator to run on that machine. That was a cool time. The emulator is all on copper squares about 1/10th inch square (the bits). That is when the term firmware really meant something because it was certainly firm.
1972 – I was doing work for a guy in Miami (Richard J. Sherin). He observed that Dike’s emulator took 3 ROM modules and each module cost around $4000. So Richard wrote a replacement emulator that only took 2 ROM modules and Richard included “hardware” floating point with his. At DNA we used that emulator until the end of the 70’s. I have one of those 30 pound ROMs in my attic. I live in FL on the beaches and it is too hot to go up there. I’ll get one down sometime and show everyone a picture of it.
Mid 70’s – Another company tried to get in on the action. The CHI 2130 (really cool and looked like a small 360). It emulated the 1130 but I think he did all of that in hardware.
Mid 70’s – General Automation had built a computer to “replace” the 1130. It was the GA 18/30. But it was all hardware. It supported lots of extra memory and large disk shared drives. I didn’t know it at the time but that is what was the basis for my emulator. GA’s customers were getting mad because GA was not delivering a multi-tasking DM2 as the promised. So GA contracted DNA to put TSO and CYTOS on it.
Mid 70’s – I met a guy at the Common conferences that worked for IBM. He had written an 1130 emulator for some IBM computer. I don’t remember which one. But it was just something to play with and I don’t know if it was full featured.
1980 – The owner of DNA, Don Nichols, got a guy at Data General to build a module for the Series/1. The board gave extra power to support TSO (our bread and butter at DNA). So Don wrote yet another 1130 emulator for the Series/1. But no one cared since all the stack based computers took the limelight. So that one became a bust.
1994 – Around 1981 one of DNA’s customers bought all of the rights to the DNA systems and got all of the source. He had General Automation 18/30’s (a mix of the 1130 and 1800). He ran these 18/30’s until I showed up at his door. He told me he had a very high electric bill because he had to run his air conditioner in the winter and he lived in KS. So being a smart #$@% I told him I would replace his power hungry computers (they were clustered) with just a few PC’s. So he contracted me to write yet another 1130 emulator. He is still running those in everyday business supporting his customers with accounting, payroll, general ledger and other things I don’t understand.
Around 1996 – I met Brian and visited his office to see real 1130’s. He showed me an emulator he wrote but I don’t know when he wrote it.
Current – lots of other emulators. Please fill this in with your dates and short history.
From: ibm...@googlegroups.com [mailto:ibm...@googlegroups.com] On Behalf Of John Doty
Sent: Friday, July 29, 2016 1:16 PM
To: ibm...@googlegroups.com
Subject: Re: [IBM1130] History of 1130 emulators
On Jul 29, 2016, at 12:13 PM, Eddy Quicksall <Ed...@Quicksall.com> wrote:
Eddy Quicksall
Eddy
To unsubscribe from this group and stop receiving emails from it, send an email to ibm1130+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
--
You received this message because you are subscribed to the Google Groups "IBM1130" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ibm1130+u...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
Al Kossow
On 7/29/16 11:29 AM, John R Pierce wrote:
> yeah, the CHI (Computer Hardware, Inc) 2130 CPU was all hard wired TTL logic. Instruction cycle times were in the
> 800nS range IIRC. I worked for CHI in 1975, and the 2130 was a done deal and had been shipped to 100(?) or so customers.
>
At the time, this was what I tracked down..
Also, CHM has some 18/30 information and card decks that I haven't gone through yet. Don't know how much GA interest
there is here.
----
Computer Hardware, Inc.
Vertical Systems
Willcox Services
--
James P. Lennane
"After receiving his master's degree in mathematics from the
University of Detroit, Mr. Lennane joined the International
Business Machines Corporation in 1964, spending two and a
half years as a scientific systems engineer and a salesman.
He then founded Computer Hardware Inc., which produced clones
of the I.B.M. 1130 and 1800 mid-size computers, and in 1973
founded System Integrators, which sold publishing systems
to small newspapers using the clone computers. The company
broke into the big time in 1980 with a $9 million sale to
The Los Angeles Times."
CHI built an expansion box for the 1130 (CHI-1114) and an
internal core memory add-on (CHI-1106).
CHI designed the 2130 and 2131, extended 1130's in '74-'75.
An expanded model (CHI-3230) was developed in the late 70's.
The assets of CHI appear to have been acquired by Vertical
Systems, Inc / Bob Willcox in the early 80's. As of 1976,
Willcox was western regional service manager for CHI. An extended
version of the 3230 with up to 2mb of memory was developed
in the early 80's.
VSI designed a follow-on system called the System 85
converting much of the discrete TTL logic to PALs and
gate arrays. It is unknown if
any of the redesigned systems were ever sold.
Willcox Services was still supporting CHI/VSI systems through
the late 80's.
Walter T. Mosscrop
My emulator is based on a Parallax Propeller microcontroller. I started in 2010, thinking that it would be a 6 month project. 6 years later, I’m still tweaking it…
I chose the “prop” because it would be a challenge to emulate an 8K words (16K bytes) machine in only 32K of main memory + 64K of EEPROM. The prop has 8 “cogs” (cpu’s), each with 2K of local storage. I emulate the 2501 and 1403 by reading and writing a microSD card. The disk is emulated by flash memory (much much faster). The blinking lights are LED’s, of course (gotta have ‘em!) and the console entry switches by “piano” dip switches. I use a 128x64 pixel dot matrix display for debugging purposes. The emulator fits on a board about 12” square.
The main reason to write the emulator was so that I could play my FORTRAN blackjack game from high school. I at first wanted to use an old DECWriter or the like… until I found out the cost! I finally settled on modifying an electronic typewriter (using another prop) as console input/output. This was not trivial… there’s no I/O interface on the typewriter. I had to tap into the keyboard control signals (info which I found on the web) to override and generate the new key scan codes that in turn allowed me to print or read the keyboard. The typewriter prints at about 9 cps, so it’s actually SLOWER than the 1130’s console printer. I miss the red/black ribbon selection, which I had used in the game for the red cards and special messages.
In all, there are 12 cogs active at the same time – 8 in the main prop and 4 in the typewriter’s prop.
During the development of the emulator, I assumed that the index registers (which of course were stored in core memory) weren’t directly accessed as memory locations. Bad assumption. The compilation of FORTRAN DATA statements, for example, does so. I spent many hours reading the assembly listings of the FORTRAN compiler’s phases.
On another note, I think it’s interesting that a 50-year-old machine has so many fans—fans that are devoted enough to create emulators, hunt down obscure bugs, or even resurrect an original machine. I also learned on a System/360, but for some reason it doesn’t have the appeal of the 1130. Maybe it was just too big?
Walter
Eddy Quicksall
It would be cool if you could emulate the Digital Scientific Meta-4. That was the first IBM 1130 clone which used ROM and firmware to emulate the 1130. I have access to the firmware listings if you wanted to try that. The Meta-4 was what is now called a RISC computer. It could hold 4 ROM modules, each weighing about 30 pounds. The firmware was etched onto fiberglass boards each about 8” by 8”. Each instruction was 32 bits. We used to “patch” the firmware by scraping off bits and pasting new ones on. Each ROM module was 4K 32 bit words. I didn’t write much of the firmware but I did write that direct disk boot code. Each memory operation was actually an I/O statement.
We used a Meta-4 from 1974 to 1982. Later DSC emulated the 360 but I don’t know how far that went.
Eddy
Al Kossow
On 8/3/16 6:18 AM, Eddy Quicksall wrote:
> I have access to the firmware listings if you wanted to try that.
Eddy Quicksall
http://bitsavers.org/pdf/digitalScientific/M4-005P-170_1130rom_Jan70.pdf
Nice. That is the listing of the DSC Firmware written by Dike Summers. It will emulate the 1130 without any frills. Richard J. Sherin wrote his own that has protected instructions and firmware floating point. We used both at DNA but sold Sherin's to those that needed to run DNA's time sharing system for the 1130 (TSO/CYTOS). I wrote a firmware XREF program for Richard when I was in the Air Force which could come in handy if he still has it.
I'll get up into the attic and find one of those ROM's and take a picture ... it is amazing. After the 1130 I worked as a BIOS engineer and Engineering Manager for AMI (American Megatrends). At AMI I kept one ROM module and one ROM board on my book case to show the young engineers when Firmware was really “firm”.
I have BCC'd Richard just in case he still has his listing.
Eddy
-----Original Message-----
From: ibm...@googlegroups.com [mailto:ibm...@googlegroups.com] On Behalf Of Al Kossow
Sent: Wednesday, August 03, 2016 11:19 AM
To: ibm...@googlegroups.com
Subject: Re: [IBM1130] History of 1130 emulators
On 8/3/16 6:18 AM, Eddy Quicksall wrote:
> I have access to the firmware listings if you wanted to try that.
http://bitsavers.org/pdf/digitalScientific/M4-005P-170_1130rom_Jan70.pdf
--
Eddy Quicksall
Here is a link to the Meta-4 firmware instruction set which was used to emulate the 1130: http://bitsavers.trailing-edge.com/pdf/digitalScientific/7032MO_Meta4Series16RefMan.pdf.
Eddy
From: ibm...@googlegroups.com [mailto:ibm...@googlegroups.com] On Behalf Of Eddy Quicksall
Sent: Wednesday, August 03, 2016 3:52 PM
To: ibm...@googlegroups.com
Walter T. Mosscrop
I reviewed the documentation for this machine. Very interesting…and fast for that time period (90 ns instruction cycle).
Any emulator for the Meta-4 would need to be implemented on something much faster than the prop. It has a 50 ns instruction cycle. So an emulation of the Meta-4 would be much slower than the real machine. On top of that running an 1130 emulator would certainly test almost anyone’s patience. The prop emulator is about 3x faster than a 3.6ms 1130.
Walter
John R Pierce
I reviewed the documentation for this machine. Very interesting…and fast for that time period (90 ns instruction cycle).
10-20 MHz was fairly common speed for microcode even back in the
mid 60s. ALL the IBM System/360's were microcoded. Main memory
was many times slower than this, of course.
The high end IBM 360 model 91 in 1967/68, had a cycle time of 60nS and could do like 16 million 32 bit integer or floating point adds per second. the 64 bit wide memory cycle time was 780 nS but it was 16 way interleaved so it could keep up with that cycle time (780/16 == 48nS). I believe a front panel from one of these is at the CHM in Mountain View.
Richard Stofer
How about using an FPGA to emulate the Meta-4 emulating the IBM1130? It is possible to get an inexpensive FPGA to run with a 10 ns clock but 20 ns is a lot easier. I haven’t really looked at the new Artix 7 chips for their maximum clock rate with my rather inelegant design. I run 20 ns on the older Spartan 3 chips, maybe with a little effort I could get the system into the 10 ns region.
eddy
I reviewed the documentation for this machine. Very interesting…and fast for that time period (90 ns instruction cycle).
Eddy Quicksall
When Sherin wrote the 1130 emulator for the Meta-4 he did not have a Meta-4 so he wrote a Meta-4 emulator that ran on our 1130 from which we ran his 1130 emulator (written in Meta-4 language). So that was an emulator running an emulator running an emulator … of the 1130 (boy that was a mouth full). We booted up DM2 and it ran so slow on the 1130 that when we typed a character at the console and it took a few seconds before it echoed back to the paper. I can’t remember how long it took to boot but I suspect we went out to get a bite that night and came back before it was done.
Boy, those were the days!!
Eddy
Paul Anagnostopoulos
I also microprogrammed an IBM 1401 emulator. It was almost impossible since the various models were incompatible. I'm not sure whether it ever became a product.
I've attached a photo of one of the ROM boards for the Meta 4. This contains 1K 16-bit words of capacitance-read ROM.
~~ Paul
Eddy Quicksall
Yea Brown! In the late 70’s I think I used the mag tape driver written by Brown. And in the mid 60’s I think we used a Fortran compiler written by Brown for the IBM 7040 … IBM’s Fortran took for the 7040 took several minutes (maybe up to 10) to compile a simple student program. All of the Fortran phases were on mag tape … it was fun to watch the 3 or 4 type drives spin back and forth. But Brown’s compiled and executed so fast (no mag tape movement) you couldn’t even read the student number as it came out on the 7040’s selectric typewriter.
Eddy
--
Eddy Quicksall
>> I've attached a photo of one of the ROM boards for the Meta 4. This contains 1K 16-bit words of capacitance-read ROM
That will save me from going up in a very hot attack and digging through lots of boxes and I’m not even sure if I’m still strong enough to pick up a module anymore … I live in Ponte Vedra Beach, FL and we really have lots of humidity and heat J. How about slipping out one of the boards and put a ruler next to it so people can see the bits?
I remembered wrong … I thought a single module was 4K but it was only 1K I guess.
I started out on the 1401. I still have the original functional characteristics manual. I’ll take a picture of part of it if anyone is curious (it is so old I am afraid of tearing the binding if I try to photo copy it).
Eddy
From: ibm...@googlegroups.com [mailto:ibm...@googlegroups.com] On Behalf Of Eddy Quicksall
Sent: Friday, August 05, 2016 8:21 AM
To: ibm...@googlegroups.com
Paul Anagnostopoulos
At some point I'll unscrew the ROM module and pull out a board.
~~ Paul
Eddy Quicksall
Oh yea, it was WATFOR for the fast Fortran compiler. That was on an IBM 7040 in 1965. Did you have something to do with that?
I remember Brown but now I think that was the tape drive routines. My customer still uses those big tapes. When I wrote that emulator I used an Overland Data board. The interface was parallel SCSI.
I already have a board pulled out. If I get to it first I’ll photo and post it. I also have a core memory board out of the Meta-4. I wish I had kept some of the wire wrap boards (about a million wires going everywhere) … people should get a kick out of that. Do you have one of those?
Eddy
From: ibm...@googlegroups.com [mailto:ibm...@googlegroups.com] On Behalf Of Paul Anagnostopoulos
Sent: Sunday, August 07, 2016 4:56 PM
To: IBM1130
Subject: [IBM1130] Re: History of 1130 emulators
Eddy, are you thinking of WATFOR 7040? If so, it was developed at the University of Waterloo.
At some point I'll unscrew the ROM module and pull out a board.
~~ Paul
--
Eddy Quicksall
I spent a lot of time on the 1130 and Meta-4 “1130” panel.
Eddy
--