• Please review our updated Terms and Rules here

Youngest 8086/8088 PC?

vitoal18t

New Member
Joined
Jun 11, 2012
Messages
7
For years, I was aware that original PC shipped in 1981 (8088) and that PC XT came around 1983.

I recently, found out that IBM's 3rd generation IBM PS/2 shipped around 1987 and that Model 25 and Model 30 both had 8086 processors offering. I believe that it has been always common to have PC for 3-4 years, so looks like 8088/8086 PCs made it well into the 1990s.

I only started using PCs in 1996, so everything before that is a "mystery" to me, only what I've been able to read from books.

So my questions is, what is the last 8086/8088 IBM PC or clone released? I'd love to hear some stories of people using 8086 machines (in office environment) in early 90s, ignoring the fact that people still use 8086/8088 PCs today in some "oddball" factory set-ups.

I guess I am just excited to find out that 8086, 286 and 386 machines were sold brand new at the same time. I guess when PS/2 Model 30 (8086) came out, both 286 and 386 PCs were already in production.

Thank you!
 
My 5160 is from 1987 - was the enhanced model with half height drives (and probably 101 key keyboard which I don't have anymore). My 5150 is from Aug 1986 but has a 1983 motherboard in it now.

I used XT clones up until about 1994 at the school I was at but they were probably made around 1988.
People often could not afford 286 or 386 offerings (here anyway). Battery powered 8088 laptops made it past 1990 from what I remember.
 
I'm pretty sure that a lot of 8088 PCs were trashed in the '90s because Windows 3.1 required a 286 and up.
 
I'm pretty sure that a lot of 8088 PCs were trashed in the '90s because Windows 3.1 required a 286 and up.
While the trashing of 8088s in the 90s may be true it wasn't because of WinBlows 3.1. A large percentage of those 286 and 386 machines you mentioned were running DOS apps and not Windows. Windows 3.1 was not all that widespread in the early '90s. And, lots of machines that did have it didn't really use it for much. They were still running the DOS apps. Remember, you still had to run DOS in order to run 3.1. So, back then, even Windows was just another DOS app. :)
 
Interesting concept. I would think the last clones would be some 3rd world clones from Taiwan or some such design. Amstrad had one at least in 1991, not sure who else but I would think perhaps mobile computing which was always behind in speed might reveal some later dates.

Interesting snippet from techrepublic:
"If the 8086 had one drawback, it was the price. The first version of the 8086 sold for $360. In 2008 dollars, that translates to over $1,200 — four times the cost of an Intel Core 2 Quad Q6700. It was so expensive that when IBM went looking for a 16-bit CPU to power its new PC, it went for the lower-cost 8088, which was introduced a year after the 8086 but used a more inexpensive 8-bit data bus."

Another interesting comment per wikipedia "NASA used original 8086 CPUs on equipment for ground-based maintenance of the Space Shuttle Discovery until the end of the space shuttle program in 2011. This decision was made to prevent software regression that might result from upgrading or from switching to imperfect clones." (the link however is only to an article from 2002 saying how NASA had to go buy them on ebay since Intel no longer supplied the chips).

Trying to come up with something here, but the Z80 sorta blows the timeline out of the water although probably isn't a valid comparison. The HP-200lx had a 80186 and was 1994.
 
I guess 8086/8088 based personal computers were made till mid-90's. Soviet ES184x (ES1840, ES1841, ES1842) computers were made at least during the first half of 90's. I believe I also saw a lot of Taiwanese Turbo XTs at that time. They were very cheap, and good enough for many application-specific uses (payroll applications, cash register, schools etc).

CMOS variants of 8086/8088 - 80C88 and 80C86 along with 82Cxx support chips are still being made by Intersil, but they are mostly used for older industrial and embedded applications:
http://www.intersil.com/content/int...nt/microprocessors-and-peripherals/80C88.html
http://www.intersil.com/content/int...nt/microprocessors-and-peripherals/80C86.html

BTW, ES1842 used a modification of 8086 called KR1810VM86M (note M suffix, 100% 8086 clone is KR1810VM86). While it supported all 8086 instructions, it generated exception on non-existing instructions (8086/8088 would just skip undefined instruction treating them as 1 or 2 byte NOPs). This allowed software-based emulation of 80286 CPU.
 
BTW, ES1842 used a modification of 8086 called KR1810VM86M (note M suffix, 100% 8086 clone is KR1810VM86). While it supported all 8086 instructions, it generated exception on non-existing instructions (8086/8088 would just skip undefined instruction treating them as 1 or 2 byte NOPs). This allowed software-based emulation of 80286 CPU.

Sergey, in the case of Intel-licensed 808x chips, that's not strictly true. Results differed by instruction and the PLA decoding. Some undefined codes simply resulted in the execution of the "closest" instruction. I think I posted about this about a year ago.
 
I think there were s-100 systems with 8085/8086 processors being made or sold in China in the 90's. Anyone know Chinese vintage computing history? The ACE Computer Enterprise exported them to China perhaps.
 
Sergey, in the case of Intel-licensed 808x chips, that's not strictly true. Results differed by instruction and the PLA decoding. Some undefined codes simply resulted in the execution of the "closest" instruction. I think I posted about this about a year ago.

Chuck, that actually is very interesting to me. I was trying to write some code to detect different varieties of 8088-compatible chips (I mean pin-compatible, 80C88, V20). And so far I didn't find any difference between 8088 and Intel 80C88. But there is a difference between Harris/Intersil 80C88 and 8088. Also it is interesting to know if there are any differences between Intel 8088 (C) '81 and Intel 8088 (C) '83. There is a detectable bug in the original 8088 with (C) 1978 marking.
Perhaps behavior of undefined instructions will be different in some of these chips.
 
Hello Everyone,

Thank you all for the information.

Personally, I still used 8086 CPU in 2004, We had to wire wrap a basic computer for our computer architecture class, when I was doing my undergrad. The class was already considered to be outdated at the time as it was getting more difficult to find 8086 and other support chips.

However, as far as Desktop PC is concerned, I think 8086/8088 machines were seeing decline in 1989 and by 1992 no one was probably buying 8086 machines as most new "Apps", like Win 3.1, Wolf 3D required 286 or higher, so financially it probably didn't make sense to invest into 8086 machine in 1992.

I have an interesting book "Beyond 640K" from 1989 and it still talks about 8086/286/386 coexisting.
I like how it talks about memory being $500 per 1MB and that it would be $8 million dollars to fill entire 4GB for 386 machine :).

Overall, it is amazing to see almost 10 years of service out of 8086, but I guess 6502 was the same way as well as other chips.
 
Computer Products United, while may not have offered an 8086/88, not sure, had an American made 80186 mobo/computer. Was offered as late as 1990. I want one. I want one now. It was kind of expensive. When I lived in Florida I called them and asked if they'd sell me a mobo, and it was in excess of 400$.
 
Why are there barely any PCs with a 186 CPU? (at least I haven't heard of any)

While the 80186 has on-board peripherals, they aren't compatible (at least not early 1980s varieties) with the standard PC stuff. As an example, consider the Tandy 2000 or the Mindset (there were others)--they each take their own version of MS-DOS.

Until it was clear that the 5150 was going to win out on the PC wars, compatibility wasn't an issue. By about 1987, however, things were pretty clear.

The 80C186EC probably comes closest to compatibility with the PC--but it's a much later chip. The NEC V40 was also a contender, used in some Japanese systems.
 
Not really sure personally, I think the 286 came out with much faster speeds and so close to the 186, and the 8088 was much cheaper that the 80186 for whatever reason ended up mostly being used in appliances and not computers.
 
Chuck, have you ever thought of authoring any computer history books? Half my posts I want to say "here's my thought, but Chuck will probably have first hand knowledge of what went on". I know you've authored software, but any books yet?
 
Years ago, I was approached by an editor from Osborne-McGraw Hill to see if I was interested (I was doing technical editing for some of John Dvrak's books at the time). So I knew what would be involved--and it looked like too much work for the money.

I'm more convinced of this now.
 
Not really sure personally, I think the 286 came out with much faster speeds and so close to the 186, and the 8088 was much cheaper that the 80186 for whatever reason ended up mostly being used in appliances and not computers.

When we were putting out our own PC, it used an 80186 and an 80286 (this was about 1982-83 and Intel was sampling pre-production samples of both chips). If memory doesn't fail me and send me digging into my files, I believe that both the initial releases of the 186 and 286 ran at the same speed--6 MHz. I also seem to recall that we were running pre-production steppings of the 80286 at 4 MHz.

Observe that the 80186 is pretty much complete in and of itself--it has its own system controller, interrupt controller, chip-select logic, DMA and timers all on the same chip. Stick on a crystal and some RAM and you've got a system of sorts. The 80826 still requires all of the external peripherals as well as a clock generator (82284) and a bus controller (82288 )--and then there's the issue of 16-bit DMA over a 16MB memory space.
 
Back
Top