From what I've been able to dig up, it's supposed to be very similar to a 386SX but was designed for embedded applications. It has an advantage over the SX in that it is capable of 26 bit memory addressing instead of 16 bit so you can have more megs of RAM than the 16 megs you would be limited to with an SX. You probably won't be able to swap it for an SX or DX because it would have to have more pins accessible by the motherboard than an SX to be able to address the maximum amount of memory. If you have a DX motherboard, you'd be better off just to use a DX. The whole point of the SX was to simplify motherboard design, so the extra pins used by the EX probably wouldn't be connected up to anything on an SX motherboard, assuming they have the same number of pins in the first place.

My EX always said I had bad memory.. ended up fighting a lot and rarely had SX. I would have preferred embedded applications on a more frequent basis..

I once found a Siemens PLC with dual 80186 processors inside. Odd thing was that they were labeled "Intel Advanced Micro Devices"

I designed that board and wrote the BIOS to it about 16 years ago. AMA. -Jim Stewart

Awesome! Maybe I don't have a good feel for the lifespan of the 386, but why was a 386 board designed in 1995?

Stull wrote:

Awesome! Maybe I don't have a good feel for the lifespan of the 386, but why was a 386 board designed in 1995?

Industrial use most likely. There's still a lot of old tech in daily use in manufacturing processes all over the world. That's one reason why old motherboards cost so much, because business users have no choice but to pay the price if they want to keep their old equipment running. It's cheaper to pay $200 for an obsolete motherboard than to pay 10's of thousands for a brand new piece of equipment.

On that note, I once read somewhere that a lot of airliners still use 386 CPUs for their onboard computer systems because they're reliable and they don't use much power. Pretty cool, eh? 🤣

I started my company with an SBC based on the NEC V25 processor. It ran DOS and had a R/W flash drive, the first as far as I can tell. A major problem with the V25 was that the on-chip peripherals were incompatible with code written for PC motherboard peripherals. The 386ex fixed that. Everything, the counter/timer, the interrupts, the serial ports were exact PC compatible (once I sorted out the chip initialization and did a proper bios). Why did it matter? Lots of customers wanted to develop in compiled Quickbasic. The compiled programs ran with no changes or device drivers on my board.

mr_bigmouth_502 wrote:

On that note, I once read somewhere that a lot of airliners still use 386 CPUs for their onboard computer systems because they're reliable and they don't use much power. Pretty cool, eh? 🤣

That's not a very comforting thought, 386's in the air traffic control towers having to make split second decisions about routing planes so they don't crash into each other (or the ground).

I didn't actually mean the control towers, I meant the planes themselves. =p

http://www.cpushack.com/space-craft-cpu.html

sliderider wrote:

mr_bigmouth_502 wrote:

On that note, I once read somewhere that a lot of airliners still use 386 CPUs for their onboard computer systems because they're reliable and they don't use much power. Pretty cool, eh? 🤣

That's not a very comforting thought, 386's in the air traffic control towers having to make split second decisions about routing planes so they don't crash into each other (or the ground).

Private pilot here.

Humans make the first-line decisions in the towers and TRACON centers. There are computers watching, but more for alerting and ratting out mistakes. Larger planes have TCAS boxes that give traffic alerts. I don't know what kind of processor they use, but the CPU load wouldn't be that great.

I took apart some LORAN receivers. They used 5mhz 8086 processors.

I think they recently "upgraded" the computer onboard Hubble Space Telescope with a radiation hardened 486...

Old tech hangs around, because it's proven tech. 😀

SquallStrife wrote:

I think they recently "upgraded" the computer onboard Hubble Space Telescope with a radiation hardened 486...

Old tech hangs around, because it's proven tech. 😀

Yea Hubble is ~ 20 years old and that was the latest technology available at that time 😀

I read an article a while ago and it mentions that hardware is one issue, but finding engineers happy to work on these (vintage) projects is another one. Many new engineers avoid such projects...

The most impressive aspect for me is the software reliability. No idea what kind of OS/software they run, but it surely is rock solid.

I'm surprised engineers wouldn't be fascinated by the longevity of the old hardware. But I suppose most people are all about the latest greatest.

Mau1wurf1977 wrote:

Yea Hubble is ~ 20 years old and that was the latest technology available at that time 😀

That upgrade was done in 1999!

Radiation hardened and MilSpec equipment lags behind desktop gear by several generations.

Yes the upgrade was done in 1999, but using old parts because that's what's compatible with the parts on the telescope.

I don't think that MilSpec was the reason for using old parts, but compatibility with existing systems.

SquallStrife wrote:

I think they recently "upgraded" the computer onboard Hubble Space Telescope with a radiation hardened 486...

Old tech hangs around, because it's proven tech. 😀

What will they do for the next upgrade or repair? Intel doesn't make the 486 anymore and there probably aren't a lot of radiation hardened versions still kicking around in the spare parts bins.

Mau1wurf1977 wrote:

I don't think that MilSpec was the reason for using old parts, but compatibility with existing systems.

Perhaps, I can't find anything about "why" they chose a 486, just that they did and "lolol old cpu"...

I wasn't meaning to say NASA would use MilSpec gear, I was just observing that those chips usually lag behind the mainstream by a fair bit.