8-core monster crunchers

_Ryle_
_Ryle_
Joined: 20 Mar 05
Posts: 8
Credit: 361505378
RAC: 7451

I always wondered why someone

I always wondered why someone like IBM or Sun doesnt enter the x86 based cpu-market. They make some powerful cpu's in their own field (i think?). I think it could be interesting with more competitors to choose from, when looking for cpus. Of course, with linux, there are distros for other architectures than x86, but for those who play games, linux still has a market to capture.

Donald A. Tevault
Donald A. Tevault
Joined: 17 Feb 06
Posts: 439
Credit: 73516529
RAC: 0

RE: I always wondered why

Message 80746 in response to message 80745

Quote:
I always wondered why someone like IBM or Sun doesnt enter the x86 based cpu-market. They make some powerful cpu's in their own field (i think?). I think it could be interesting with more competitors to choose from, when looking for cpus. Of course, with linux, there are distros for other architectures than x86, but for those who play games, linux still has a market to capture.

That's mainly because of licensing issues. The only reason that AMD can manufacture CPU's with the x86 instruction set is because Intel signed a cross-licensing agreement with them back in the 386/486 days, so that AMD could be a second source of x86 processors. (That was back when Intel didn't have sufficient manufacturing facilities to meet demand, and before they realized that AMD would someday become a competitor.)

IBM did manufacture x86 processors for a while, several years ago. (That was their "Blue Lightening" line of 486 processors.) I believe that they also had a cross-licensing agreement with Intel, but I don't know what became of that.

archae86
archae86
Joined: 6 Dec 05
Posts: 3145
Credit: 7023824931
RAC: 1805761

RE: RE: I always wondered

Message 80747 in response to message 80746

Quote:
Quote:
I always wondered why someone like IBM or Sun doesnt enter the x86 based cpu-market. They make some powerful cpu's in their own field (i think?). I think it could be interesting with more competitors to choose from, when looking for cpus. Of course, with linux, there are distros for other architectures than x86, but for those who play games, linux still has a market to capture.

That's mainly because of licensing issues. The only reason that AMD can manufacture CPU's with the x86 instruction set is because Intel signed a cross-licensing agreement with them back in the 386/486 days, so that AMD could be a second source of x86 processors. (That was back when Intel didn't have sufficient manufacturing facilities to meet demand, and before they realized that AMD would someday become a competitor.)

IBM did manufacture x86 processors for a while, several years ago. (That was their "Blue Lightening" line of 486 processors.) I believe that they also had a cross-licensing agreement with Intel, but I don't know what became of that.

Not quite.

Intel did license AMD to manufacture the 8086. We even transferred production tooling to them. We were not motivated by manufacturing capacity, but by customers who wished to have an assured second source. Lots of other folks made 8086s without license. I think I personally saw about half a dozen, and doubt I saw them all. Even the Russians made one, and I can assure you that they did not request a license.

The big change came with the 80386, and the sign was opposite that you have indicated. That was the generation for which Intel ceased inviting anyone to make our microprocessors. I heard that was Andy Grove's personal decisions, but can't vouch for that point.

Even within the last few years there have been x86 designs from folks other than AMD. Via for a while had them from two completely different sources, Cyrix and Centaur. NexGen, Chips and Technologies, and Transmeta all had a go.

There are legal obstacles, but they are not impermeable. In the Cyrix example, mutual claims of patent infringement were resolved by letting each firm manufacture. So that type of cross-license arising out of legal proceedings continued.

Making an x86 CPU design that both works and is adequately compatible is actually quite difficult. Making a viable business of it is even harder.

peanut
peanut
Joined: 4 May 07
Posts: 162
Credit: 9644812
RAC: 0

A bit off topic but I saw

A bit off topic but I saw reference to 8086 below. The place I work has equipment controllers powered by 8086s. I even burned eproms for the boards when making s/w mods. Those 8086s take a licking and keep on ticking!

archae86
archae86
Joined: 6 Dec 05
Posts: 3145
Credit: 7023824931
RAC: 1805761

RE: A bit off topic but I

Message 80749 in response to message 80748

Quote:
A bit off topic but I saw reference to 8086 below. The place I work has equipment controllers powered by 8086s. I even burned eproms for the boards when making s/w mods. Those 8086s take a licking and keep on ticking!

Thanks for the post. The world's first 8086 die came to life on my wafer prober about September 1977. First production shipments were in June 1978. Nice to hear direct testimony that some are still around. That September 1977 stepping had fatal flaws and a pinout different from the shipped version, so the ones where you work are at the very least from a version a few months later, and most likely later than that still.

Mike Hewson
Mike Hewson
Moderator
Joined: 1 Dec 05
Posts: 6534
Credit: 284710168
RAC: 110462

RE: Thanks for the post.

Message 80750 in response to message 80749

Quote:
Thanks for the post. The world's first 8086 die came to life on my wafer prober about September 1977. First production shipments were in June 1978. Nice to hear direct testimony that some are still around. That September 1977 stepping had fatal flaws and a pinout different from the shipped version, so the ones where you work are at the very least from a version a few months later, and most likely later than that still.


Actually I think it is soooo cooool that you contribute here, an 'obstetrician' of the '86 line ..... ah, what can I say but 'segment:offset' ..... :-)
I still remember my first 8088 that I cut my 'assembly teeth' on, just after I put my beloved Commodore 64 away [ & ever thankful to move away from the punch-card queue at the PDP-11 terminal in uni ].

Cheers, Mike.

( edit ) I've still got this book called the 'Commodore 64 Whole Memory Guide' .... probably the last book of it's type ie. one that could discuss an entire machine's memory space in such detail.

I have made this letter longer than usual because I lack the time to make it shorter ...

... and my other CPU is a Ryzen 5950X :-) Blaise Pascal

Winterknight
Winterknight
Joined: 4 Jun 05
Posts: 1220
Credit: 312182938
RAC: 656503

My firm still supports

My firm still supports industrial equipment control and monitoring modules based on Z80 cpu, although these are being phased out. The replacement is 386 @ 10MHz, vastly over specified for job and there are no plans, as yet, for that to be redesigned.

Mike Hewson
Mike Hewson
Moderator
Joined: 1 Dec 05
Posts: 6534
Credit: 284710168
RAC: 110462

RE: My firm still supports

Message 80752 in response to message 80751

Quote:
My firm still supports industrial equipment control and monitoring modules based on Z80 cpu, although these are being phased out. The replacement is 386 @ 10MHz, vastly over specified for job and there are no plans, as yet, for that to be redesigned.


But a couple of years ago the Shuttle people were calling for 486's - for their compatibility largely, but also the transistor size is reasonably resistant to radiation in a way that current designs aren't. Presumably the smaller resolution of more modern components make it easier to knock them out with the odd high energy cosmic ray etc ...

Cheers, Mike.

I have made this letter longer than usual because I lack the time to make it shorter ...

... and my other CPU is a Ryzen 5950X :-) Blaise Pascal

Winterknight
Winterknight
Joined: 4 Jun 05
Posts: 1220
Credit: 312182938
RAC: 656503

RE: RE: My firm still

Message 80753 in response to message 80752

Quote:
Quote:
My firm still supports industrial equipment control and monitoring modules based on Z80 cpu, although these are being phased out. The replacement is 386 @ 10MHz, vastly over specified for job and there are no plans, as yet, for that to be redesigned.

But a couple of years ago the Shuttle people were calling for 486's - for their compatibility largely, but also the transistor size is reasonably resistant to radiation in a way that current designs aren't. Presumably the smaller resolution of more modern components make it easier to knock them out with the odd high energy cosmic ray etc ...

Cheers, Mike.


I happen to know where there are quite a few 486's. Probably several 100, another industrial control and measurement system that is being re-furbished. I'll talk to my friend to see what the score is, before they throw them out. He said about two weeks ago, at the moment, the computers are been kept as spares until re-furbishment is completed.

tullio
tullio
Joined: 22 Jan 05
Posts: 2118
Credit: 61407735
RAC: 0

RE: RE: Thanks for the

Message 80754 in response to message 80750

Quote:
Quote:
Thanks for the post. The world's first 8086 die came to life on my wafer prober about September 1977. First production shipments were in June 1978. Nice to hear direct testimony that some are still around. That September 1977 stepping had fatal flaws and a pinout different from the shipped version, so the ones where you work are at the very least from a version a few months later, and most likely later than that still.

Actually I think it is soooo cooool that you contribute here, an 'obstetrician' of the '86 line ..... ah, what can I say but 'segment:offset' ..... :-)
I still remember my first 8088 that I cut my 'assembly teeth' on, just after I put my beloved Commodore 64 away [ & ever thankful to move away from the punch-card queue at the PDP-11 terminal in uni ].

Cheers, Mike.

( edit ) I've still got this book called the 'Commodore 64 Whole Memory Guide' .... probably the last book of it's type ie. one that could discuss an entire machine's memory space in such detail.


Having worked at SGS, once Olivetti Fairchild and now STMicroelectronics, in Agrate Brianza on the Z80 and Z8000 with people that had been colleagues of Federico Faggin I still have a copy of the 8080 Microcomputer Systems User's Manual, September 1975. Manuals were very different then.
Tullio

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.