R15 computer

Since the machines of the ESZR -I series could not catch up, the ESZR-II series was launched, to which Hungary joined with the R15 computer developed at SZKI.

Unfortunately, no evaluable documentation has survived from the computer, which is why the product sheet is incomplete. The data provided below is mostly derived from verbal communication from those involved in the development.

Category:
Medium machine
Technology:
Custom semiconductor
Origin:
domestic
Creation:

The aim of the development was to meet the basic objectives of the ESZR-II series: at least instruction-level compatibility with the IBM 370 machine family, as well as significantly greater main and secondary storage capacity and computational accuracy.

Developer:
SZKI Hardware Systems Engineering Laboratory (HRL)
Designer:
Led by Pál Németh - Chief Designer
HRL development metrics
Manufacturer:
Computer Science Experimental Plant Limited Liability Company (SZKÜBT)

Use

Operating time:
Issue: 1978-79
Working quantity:
11
Typical applications:

Corporate data processing

Operating locations:

Large state-owned enterprises:

  • TRY
  • ETERNAL
  • Budapest Dairy Company
  • HEAT ENGINE

 

Structure

Central unit

Byte-organized machine (8 bpB)

  • controller: control circuits built from Schottky TTL (STTL) elements
  • main memory: 64 KB ferrite ring RAM
  • calculator: decimal arithmetic built from TTL elements
  • channel: IBM compatible multiplex channel

Periphery

  • storage: up to 4 EC 5052 (Bulgarian) hard drives
    • 7.25 MB removable disk pack
  • external devices:
    • operator desk (VT340 monitor + console typewriter)
    • console matrix printer (LX 120)
    • punched tape reader (MOM)
    • punch tape punch (MOM)
    • line printer ( VIDEOTON )

Operation

The machine worked with 16 and 32-bit, single- and dual-address instructions, with an 8-bit opcode. The developers implemented the full instruction set of the 370/115 model, which - because some instructions had multiple versions thanks to the 4 mask bits - included much more than 256 instructions.

Main groups of instructions:

  • data transfer within the main memory and between the main memory and registers
  • data transfer between main storage and external devices (in/out operations)
  • decimal arithmetic operations (with EBCDIC encoded decimal numbers)
  • logical operations (comparisons, shifts)
  • control instructions

There is no evaluable data on execution times.

Program set

General programs
  • operating system: "English" IBM DOS
  • compilers: na
Custom programs
n.a.

Historical curiosities

The R15 was a true ESZR machine: the goal was to create a machine with essentially the same architecture and instruction set as the prototype (IBM 370/115). However, its reliable operation could not be achieved: the mean time between failures (MTBF) was 20–30 hours.

The great development engineer Endre remembers the reasons as follows:

“Technological problems were insurmountable obstacles.

First, all the logic circuits had to be redesigned from ECL to the completely different TTL element set. This Sisyphean task, which was ultimately successful, took about a full year.

However, the available Schottky TTL element set had an unpleasant property that aggressive "spikes" were generated at the output of the chips, and if these exceeded the permissible noise level, they caused incorrect switching, and the operation became uncontrollable.

The difficulties were compounded by the fact that only two-layer PCB technology was available, where – in the absence of a separate power supply plane – it was not possible to provide a stable power supply to the chips.

The service was based on a very advanced diagnostic system, which stopped the machine in case of an error. If the damaged data could be restored based on the error messages, the service processor automatically continued the interrupted program, but otherwise all partial results were lost and the program had to be restarted from scratch. It may not seem immodest to say that I wrote the entire diagnostic software myself, it was my last work on the R15 and it took about a year.

By the time these problems became apparent to the decision-makers, nothing could be done without the entire program collapsing. So we finally completed the development – with great difficulty – and the machine was approved. But the reliability of the R15 did not reach the level that could be minimally accepted from a machine that would be put into normal use.”

The end of the story is that SZKI had 11 copies manufactured by SZKÜBT and these were given to large state-owned companies - presumably with the cooperation of OMFB.

Resources

Oral statement by development engineer Endre Nagy

General description of the IBM 370 system (English)

IBM 370 System Operating Instructions (English)

 

 

 

 


Created: 2018.10.22. 13:57
Last modified: 2022.04.14. 19:04
Translation

× Close