Personal Computer Museum, Canada's Videogame Museum

GenRad Futuredata 2300

GenRad Futuredata 2300

Speed2 MHz
Memory48 KB

What's this?

Release Date: 10/1/1979
Manufacturer: GenRad
Original Retail Price:
Adjusted Inflation Price:
Donated By: Rick Hille
This system is actually primarily a development system. It contains:
  • Intel 8080 processor 2 MHz.
  • Built-in terminal, monochrome, green.
  • S-100 bus interface (common at the time)
  • In Circuit Emulator modules (options) for 8008, 8080, 6502, and 6800 CPU's
  • Casette tape drive (option)
GenRad Futuredata 2300 Rear
John Oldenkamp discusses his system:

We bought the FutureData in early 1980 and at that time, it was a recent demo unit so I would think 1979 would be about right for its release date. GenRad (General Radio) had some earlier models and we bought the latest and greatest. We paid about $21K for it, equipped with 2x 8" Shugart FDD (2x250K), 48K of target memory (expensive option), and the in-circuit emulator/logic analyzer module for an 8085 target. We considered the full 64K option but much of that was taken up by the OS, video etc.

Like most development systems of the day, it ran with the native processor i.e. if you were developing 8085 code, you bought an 8085 CPU for the FutureData, for a 6800, the CPU was a 6800 etc. They supported most of the popular 8 bit processors (we had a Z80 CPU board but never used it). The CPU had to match the target because the in-circuit-emulator ran inside the FutureData itself alongside the OS. We used the logic analyzer and emulator infrequently, preferring to download the code to a custom monitor program on the target but it worked and was pretty zowie for the day.

Microcontrollers were supported with cross assemblers (we had the 8048 one) but did not have ICE support since the microcontrollers of the day could not run the FutureData OS.

The oscillator was (I think) socketed so that you could match the emulator speed to your target. The entire system ran at that rate. It wasn't much, ours ran at something less than 4MHz. About average

User Comments
Thad on Saturday, April 28, 2018
I had several GenRad 2300 ADS systems back in the day. I ran a BBS on it for years running under CP/M. Oh how I miss the 8" disks and the massive "card cage" in the rear. Great memories and a great machine at the time!
Michael Fleming on Thursday, May 21, 2015
Memories :-) I used the Z80 version and developed a Digital Tracking Filter for Boeing in 1981 that became stand equipment on all 767's and 757's -- as a development system it had all I needed -- I really liked the split screen debugger, very advanced for its day.
Craig Landrum on Friday, January 9, 2015
By the way, these units used an early capacitance-based Keytronics keyboard, and after several decades, the foam disks in the keys tend to rot away. I completely revamped mine at home with new foam disks and should anyone be interested, I can supply a document that describes the step-by-step procedure to restore the keyboard. Contact me at craigl at
Craig Landrum on Friday, January 9, 2015
I was with Planning Research Corporation from 1978 to 1988 and we used GenRad Futuredata 2300 systems to develop all of our Z-80 based firmware for the custom multibus based boards we developed at the time for use in our pioneering ImageNet document scanning systems. The success we had with our early adopters allow PRC to win the large US Patent and Trademark automation contract in 1984. When I left PRC in 1988 to found my own document management company, they let me take the GenRad systems with me since they had no projects left that needed Z-80 firmware development. I still have two systems, both of which are in working condition (in 2014), but only one has a complete set of memory boards, the other has none (any memory boards I could obtain would be most welcome!). I can contacted at craigl at
Andy Patterson on Wednesday, July 30, 2014
I started working for FutureData about the time they were purchased by GenRad. They were sold to Kontron. The whole time a was there they used a z80 processer in the development computer and a seperate target processer in the In Circuit Emulator probe. I was part of language development group and became the head of the language development department at Kontron. How the ICE units worked depended on the target processor.
Phil Clements on Friday, September 27, 2013
I was involved in sales and support of the Futuredata AMDS in the UK. We first got involved with Futuredata around 1977/78 and their first system was smaller with a 14 inch TV as a monitor. It supported the 8080, 8085 and Z80 and had the assembler, linker and debugger held in RAM. At the time it was quite advanced compared with any competition. Unfortunately, I have lost my data sheets for the product, but hopefully they may still be hiding in some corner of my garage? My main memory of Futuredata was all the promises they made about future hardware and support, which encouraged GenRad to purchase the company. As it turned out, GenRad were sold a pup and it took them two or three years to discover it. I will continue looking for my info and post any further info.
David Stonier-Gibson on Friday, May 25, 2012
A correction on the AMDS. It was not an S100 bus. It just happened to use the same edge connector. I was the Australian sales and support engineer for FutureData in 1979/1980, working for Promicro Pty Ltd, a subsidiary of Datacraft. I had one of the very first AMDS units built, and provided better support back to the factory than they provided me! The unit I had cost about 2 years salary for an experienced electronics engineer (me!). It had a 4MHz Z80, 64K static memory (dynamic was cheaper) and an all-important in circuit emulator. The emulator was actually an interface off the (non-S100) bus, which itself reflected the CPU chip via buffers. It used one CPU to run everything. The debugger had to share memory with the target application. Everything was done in assembler. There was a single hardware breakpoint, implemented as a bunch of TTL chips on the debug/emulator board. It pulled a non-maskable interrupt. The debugger software would determine if the NMI source was the breakpoint hardware or the target system, and branch accordingly. Thus NMI never gave true timing. Despite such shortcomings, the AMDS was the very best universal development system of its day. With different CPU boards it could support 6800, 8085 and 1802. Unfortunately FutureData could not withstand the competition from big name companies. Tektronix has a system. It was an absolute dog, but got sales because of the name. HP came out with a system at about 2-3 times the prices, and cleaned up with all the government buyers. I still have an AMDS under the stairs at work. It got me started in my own business in the early 80's and was still in use until about 1989. David Stonier-Gibson
Paul Raveling on Friday, October 16, 2009
In the preceding comment I forgot to mention the Intel 8088, 80186, and 80286 as more chips supported by the 2302 product line by 1983.
Paul Raveling on Friday, October 16, 2009
This example is very slightly earlier than 1979 in terms of what we used in-house at that time. I joined FutureData in 1979 and became manager of software development for the 2302 Slave Emulator product line, which involved writing a new debugger for what was called the ADS 2 (Advanced Development System) at the time I joined the company. The ADS 2 and each 2302 emulator box used a nearly identical hardware configuration with a 4 MHz Z-80. In that generation of the 2300 the tape i/o was replaced by 8-inch floppy disks. The 2302 was a dual-processor system, split into the IP and the EP -- Interface Processor, with the Z-80, and Execution Processor, using the processor chip to be emulated. The emulator's first target chip was the Intel 8086. Other target chips added to the product line while I was there were Intel 8080/8085, 8048, and 8051; Motorola 68000, 68010, 68020, and 6809; Zilog Z-8001, Z-8002, and Z-80; MOSTEK 6502.
Have a comment about this Computer (personal stories, additional information)? Post it here (no registration required).

Share |

* Inflation data courtesy of Values are approximate using our own calculations.