Talk:Intel 80286

From Wikipedia, the free encyclopedia

and could run up to 12.5MHz.

I removed this, because I think that it is wrong, but I'm not sure. I believe that a 16MHz 80286 was pretty common, and my father said that his first computer had a 20MHz 286 CPU. Can anyone confirm or deny? -- Stephen Gilbert

UPDATE: I am sure. Check this page: http://www.pcguide.com/ref/cpu/fam/g2I80286-c.html -- Stephen Gilbert

It seems there were 25MHz 286s too, though not from Intel. Crusadeonilliteracy


codenamed iAPX

I removed this because the 80286 was not codenamed iAPX. The iAPX was a completely different processor family developed by Intel in 1981. At first the family consisted of a two-chip general data processor, and a single-chip interface processor. Eventually Intel added a bus interface unit and a memory control unit. Please see a great iAPX website at http://www.brouhaha.com/~eric/retrocomputing/intel/iapx432/

Chris Thames mavroxur@hotmail.com

Contents

[edit] Exiting protected mode

However, the 286 couldn't revert to real mode, so protected mode wasn't widely used until the appearance of the 386, which could go back and forth between modes.

It should be noted that although this is technically true, IBM and others created workarounds for this limitation early on, so it's not as big a problem as the article makes it sound.

References: Protected mode overview Triple-fault technique

Tim Peterson 18:03, 22 Apr 2005 (UTC)

[edit] The 80286's performance is more than twice that of its predecessors (the Intel 8086 and Intel 80186)

You know that section is pretty confused: The 80286 was perhaps twice as fast as it's predecessor, the 80186, because it typically ran at a higher clock speed. But it was much faster than the computer most people had, because like in the 80186, calculation of the more complex addressing modes (such as [BX+SI]) were performed by dedicated hardware; the widely used 8088 had microcoded intstructions (like a RISC processor) that took many clock cycles to complete. In fact, the performance increase per clock cycle between the 80186 and the 8086/8088 may be the largest among the generations of x86 processors. In comparison, (and speaking as one who used these processors) the step between the 80186 and the 80286 was negligable. On the other hand, I hate to say that you are just wrong: most people went straight from an 8MHz 8088 to a 16MHz 80286, and they were looking at a massive difference.


I looked at the 8086 and 80386 instruction manuals, and it seems to me that there is a difference, but i dont see the doubling in many common instructions. It might be true that all the complicated memory addressing is much slower, but for example add went from 3/10 to 2/7. It's true that push/pop are much faster etc, but i guess there is not much more than a doubling of performance per cycle overall. The jump from 386 to 486 saw an approximate doubling in performance, too. so i dont understand this statement. anton
After review the manuals, I noticed that I was using an 80186 manual. The 80186 is probably closer to the 80286 in terms of performance than the 8086. anton

[edit] Possible err in para 1

0.21 million instructions per clock? (first paragraph). Surely not? prehaps 0.21 MIPS, but not 0.21 M IPC?!. (Sam Cadby) —Preceding unsigned comment added by 217.33.137.68 (talk) 16:47, 26 September 2007 (UTC)

[edit] 4 MHz 286 chip?

I was wondering if someone should write something about the Intel C80286-4 in the article. The processor's maximum clock speed should be 4 MHz, and was for a short period manufactured and used in some personal computers, for example, the Ericsson 286 (which I personally owned once). The processor has LCC packaging. There seems to be conflicting information about this processor on the internet, and rarely have any information on the LCC packaged processor. Not sure about the speed though, maybe someone knows a bit better?
See more information at [1]. --62.142.251.79 (talk) 01:50, 21 February 2008 (UTC)