Restoring a Texas Instruments 99/4A Home Computer

In October 1979, semiconductor giant Texas Instruments entered the home computer market with its TI-99/4 all-in-one computer, announced at a retail price point of US$1150 (around US$4400 in March 2022 BidenBucks). Unlike competitors from companies such as Commodore, Apple, and Tandy/Radio Shack, who used 8-bit microprocessors such as the 6502 and Z-80, the 99/4 employed Texas Instruments’ TMS9900 16-bit processor, which implemented the instruction set of the company’s minicomputer line and was also the processor used in the Marinchip Systems M9900 machines for the S-100 bus which I had introduced in 1978.

The TI-99/4 was a fine example of how Texas Instruments, an engineering company par excellence, was blind, deaf, and all thumbs when it came to marketing. The machine, equipped with a 16-bit processor somewhat similar to Digital Equipment’s PDP-11 in architecture, was equipped with only 256 bytes of 16-bit memory and 16 Kb of 8-bit video memory which the processor could only access indirectly and slowly. It was intended to run canned applications from cartridges, and no information was provided for independent developers. Jerry Pournelle wrote at the time “TI’s message is loud and clear: 'Drop dead, hobbyists!”. The machine came with a calculator-style membrane keyboard which, along with its nonstandard layout; made typing difficult. Its only programming language was TI BASIC, which benchmarks showed to run around half the speed of Microsoft BASIC on pure 8-bit machines.

In June 1981 an update, the TI-99/4A was announced at a lower price point of US$525 (US$2000 today) with additional peripherals available (most costing around twice competitors’ equivalents) and a real keyboard, but retaining the closed architecture.

This was a colossal flop, and the series was discontinued in March 1984 after racking up losses over its history estimated at around half a billion then-dollars. After the product was discontinued, the remaining stock was sold out at US$49 for the complete computer.

This reminds me that as I am exposed to the varied aspects of modern computing, most every time, I have regret about my inability to understand. I think it is, indeed, inability; that no amount of reading or study (only partly due to my age) could equip me with the profound understanding I wish I had. My sense is that one must have ‘lived’ the development as John Walker has, in order to have experienced the ‘happening world’ of computing as it unfolded, the ‘context’ of the events, what the events ‘continuity’ entailed, and how individual developmental episodes enabled ‘tracking with closeups’. i.e. to have “Scanalyzed” the field of computing and networking to arrive at a deep understanding.

This view that having ‘lived’ the development of modern computing, derives from my own experience with learning medicine. The field is so broad, I know only a little of most specialties beyond my own. However, medical school taught me the vocabulary and language of medicine. It also immersed me in the intellectual history of how knowledge was acquired in most all the important areas of medicine. My acquaintance is lifelong and the sheer quantity of knowledge in most every field is so extensive in every field that few individuals can be truly knowledgeable in more than one single field. True polymaths are rare. Given the exponential growth of knowledge and linear capabilities of the (un-augmented) human brain, there are likely to be even fewer in the future.

1 Like