socha@drivax.UUCP (Henri J. Socha (7-x6628)) (12/17/87)
In article <430@gethen.UUCP> farren@gethen.UUCP (Michael J. Farren) writes: >In article <2804@drivax.UUCP> socha@drivax.UUCP (Henri J. Socha (7-x6628)) writes: >>8080 was originally based on the Intel 8008. >>working. It was called the Datapoint 2200. Introduced in '71 I think. > >Well, you got the name of the company Intel designed their first micro- >processor for right, anyway. It wasn't the 8008, though. That was Intel's >own design, done after they saw some of the potential inherent in a >general-purpose controller design. The design for Datapoint, which piqued >Intel's interest, was the 4004, an entirely different kettle of fish than >the 8008. And saying that the 8080, and, through it, the 80386, was 'the >same' as the 8008 is rather like saying that a Vax is 'the same' as the >original PDP-11. Obviously a relative, but hardly equivalent. WRONG! WRONG! WRONG! To quote from Robert Noyce Vice Chairman, Intel Corp. in E.E. Times Aniversary Issue Nov. 87 Page A15 "It all started in November 1971, when Intel introduced its 4004, the first commercially successful microprocessor. This product was originally a response to a request from Busicom, a now-defunct Japanese calculator manufacturer, ..." "While this microprocessor [4004] was being developed, work was also begun by Intel and Texas Instruments on 8-bit CPUs for Computer Terminals Corp. (now Datapoint). Though CTC decided not to incorporate either company's product..." [because they had is 1 year earlier in MSI] So, your first points are now refuted by someone of authority. Now, for the second: that the 8008 is not the same as the 80386, I agree in general but, the ancestry is visible. As an example I once had to write a test programme for a removable hard disk drive for a Z80 based machine. Now, I know the Datapoint architecture machines better so decided to write/debug it for that one instead. The programme was in assembly language and I was able to cross-assemble! to either instruction set. Yes, there were places in the code where I had to know which machine I was running on but this was OS related and not instruction set (processor) related. There is a common set of op-codes (most of the instructions) which is common between the Datapoint 1800 (a successor to the Datapoint 2200) and the Z80 (outgrowth of the 8080). The difference is ONLY the bit patterns not the functionality of the instructions! The Datapoint 1800 had to be op-code pattern compatible with the 2200 which is EXACTLY the op-code pattern of the 8008! Now, from the 8080 (Z80) to the 8086 is a known step. I think Intel even had a translator programme. And the NEC V20 & V30 can do both because they are so similar). In fact I have heard said that the IBM-PC's BASIC in ROM is just a cross-assembled 8080 assembler! And, from the 8086 to the 80386 are also known steps. That is why the Segement register got renamed the Selector register, just an extension of the original idea. PDP-11 --> VAX is actually much of a greater leap than 8008 --> 80386 but yes, there are great differences. -- UUCP:...!amdahl!drivax!socha WAT Iron'75 "Everything should be made as simple as possible but not simpler." A. Einstein