Talk:TI-990
This article has not yet been rated on Wikipedia's content assessment scale. |
How many bits wide?
The main page of this article could be improved by saying how many bits wide each register was.
If this microprocessor is 64K wide, it would be safe to say that the program counter was 16 bits wide, but would that be true of all the other registers if this were a truly orthogonal set of instructions? Dexter Nextnumber (talk) 04:25, 7 May 2010 (UTC)
Registers at coldstart?
[edit]Can anybody describe what happens to the "registers" (the first 16 bytes of addressable memory, apparently) - are they cleared to zero? How is the program counter loaded with an initial value?
How were interrupts handled by the TI-990? Dexter Nextnumber (talk) 06:45, 18 May 2010 (UTC)
floating point
[edit]Anyone know how floating point is done? It seems that there is an option for a hardware floating point unit, or software emulation of one. Gah4 (talk) 12:15, 25 June 2021 (UTC)
Too technical
[edit]I've been editing this for clarity. Then working on a few details. But:
The article is all technical. It reads like the assembly language programmer's manual.
Ideas:
What was the story of this machine? Who designed it? (research: Harvey Cragon) Where was it made, sales and marketing, competition. How did Texas Instruments rely on it internally (see IEEE articles.) Who typically bought it, Small business, Scientists, Industrial? (see trade mags) How was it bundled? Typical peripherals?
What was the software situation? Third parties? One source will be TIMIX newsletters.
Reception: how was the 990 reviewed? Cite Mini-Micro Systems, Electronics Magazine, and others. Show comparative sales numbers.
What about AMPL 990 systems?
When did the 990 series end?
What was innovative? What patents did Texas Instruments file?
The 990/10A story is unique. The twin purpose of the 99000 chip : update the 990/10, and one last attempt to boost the 9900 to compete with 8086, 68K and Z8000. (There, the 9900 gives way to the DSP story.) Tension between DSG asking MOS for a new minicomputer chip. Divisions at Texas Instruments were more competitive than cooperative.
The last hurrah for unique aspects of the 990 was the TMS380C25 chip in 1995.
Technical again: the instruction groups could be presented in less space and with the names for the groups. Jargon should conform to that manual or be widely recognized.