Today, SoCs (System on a Chip) have become common in all kinds of devices, from smartphones to televisions, video game consoles or computers. The industry seems very clear that the future is to have the more the better within the same chip. But it wasn’t always like that.

Our story begins at the beginning 80’s. Personal computers were taking their first steps, and engineers ran into a big problem: the technology of the day it is not possible to increase the number of transistors chip with the required speed and CISC designs stagnated, so many went to parallel processing: divide the work between multiple processors was the most viable option and everyone was convinced of it the software of the future will be ready for that.

Almost as a surprise, Inmos, a small British company based in Bristol, surprised the world with the first universal microprocessor designed for parallel processing: converter. David May Y Robert Milne They were the ones who made the miracle possible.

The name transputer combined concepts transistor and computerclarifying its purpose: just as transistors make up a processor, transputers would be the basic unit, the “blocks” with which build very powerful computers at low cost. Parallel computing would do wonders.

Inmos’ aim was to create a catalog of transputers with different features and prices, which included: on the same plate silicon, one PROCESSOR, Memory FRAMEsupport for buscommunication series and even his own operating system real-time (RTOS). To make them work, it was only necessary to provide them with electrical power and a clock signal to coordinate operations.

That is, only for the price a few dollars per unitTransputers could be used as the main CPU (or part of it), but also as a disk controller or for communication control. While in traditional systems a processor dedicated to a disk controller, for example, is idle when not being accessed in a transputer system, those processor cycles can be spent on other tasks, dramatically increasing the overall performance of the machine.

Already in the basic version each transputer could connect with four otherswith bandwidth up to 20 Mbit/s (very high for those years). Each group, in turn, could connect with the others, in connections of up to tens of meters, and thus build powerful computing farms.

Transputers connected in a fixed point-to-point pattern, so communication between two distant ones required passing through all intermediaries. This caused delays in large networks, which Inmos sought to solve by integrating a zero-delay switch capable of connecting up to 32 transputers simultaneously.

“Transputers” were designed to be programmed with the Occam language developed by David May and are based on the computation of sequential processes. Inspired by CAR Hoare’s CSP (Communicating Sequential Process) paradigm, it supported concurrency and communication between processors. Subsequently, implementations of more conventional programming languages ​​such as C, FORTRAN or Pascal were released.

Inmos announced the launch of the first transputers in 1983, although they were not available until the following year (for reference, the original Apple Macintosh was also released that year). They contained 2KB of RAM and an integrated memory controller so that more could be added without the need for additional hardware. 16-bit T2, 32-bit T4 and 32-bit T8 series were released with 64-bit floating point support.

converters virtual memory support is missing, which prevents the adaptation of variants of Unix operating systems that started to become popular (variants were modified, e.g. Minix or Idris). Commercially, Inmos remained in no man’s land: Although workstation performance was very good at the time of release, there were already RISC alternatives that surpassed it. In personal computers, where 8-bit and 16-bit machines dominated the fiercely competitive arena, any Inmos option was too powerful and expensive for users.

Even the supposed advantage of building machines by combining blocks disappeared when the T800 hit the market $4,400 each, making any alternative comparison of price and performance impossible. The Atari Transputer Workstation, released in 1989 and based on the T800, is a classic transputer-based commercial example.

Inmos had time to make one more attempt: T9000 it had many features in common with the T800 with the addition of support superscalarity. It featured true high-speed 16KB cache instead of RAM and significant bus improvements. However, the funds ran out, sales of the previous models did not go well and finally the company acquired SGS who eliminated the T9000 and opted for less innovative projects, but more profitable as well as the design of microcontrollers for consumer electronics.

It was considered a transputer for several years in the late 1980s an architecture that would mark the future of computing. Neither the Inmos nor the transputer failed, but we have to admit that influenced the development of new ideas, several of them are part of systems we are all familiar with.

More information:

https://www.atarimagazines.com/startv4n4/transputer.html

http://people.cs.bris.ac.uk/~dave/index.html

https://www.transputer.net/tn/06/tn06.html#x1-20001