This article gets some things right but it gets plenty of them wrong. In particular, " Technically speaking, the industry is mired in hardware standards (Intel and Motorola CISC processors) with growth rates that are flattening out relative to the state of the art - just as the 360/3090 and VAX architectures did."
The Motorola 68k line really did reach the end of its rope in the 1990s and everybody from Sun to Apple to Amiga had to plan an exit strategy or go out of business.
The Intel x86, on the other hand, continues to outpace everything else on practical performance. ARM is coming on strong in tablets and smartphones, but overall, the x86 platform has lasted a long time.
Not long after he wrote this, IBM's System/360 architecture went through an amazing transformation that's kept it competitive.
From the beginning, the System/360 was based on bipolar logic that consumes much more power than the CMOS logic microprocessors are based on. In the early 90's, IBM realized that further development of bipolar logic was impossible because of power and heat dissipation issues.
IBM had to go to CMOS, but CMOS logic wasn't as fast as bipolar logic at the time, leaving IBM unable to produce mainframes as powerful as the last generation -- unable to handle the requirements of existing customers.
IBM's answer was "Parallel Sysplex", a clustering solution that presents a single-system image of a group of mainframes. With Parallel Sysplex, IBM produced microprocessor-based mainframe clusters that scaled beyond the last generation of bipolar mainframes.
Some architectures end up being dead ends, but if something is commercially successful, it's amazing how long it can sometimes be strung along.
Just curious: if the 68k series had had the same amount of resources put behind it as x86, there's no reason to think it wouldn't be similarly competitive today, is there? (Obviously this was historically impossible because nobody had the incentive that Intel/AMD have had, so I'm just asking from a technical standpoint.)
They expose the x86 instruction set externally for compatibility, but translate it to a different instruction set internally. It is conceivable that you could have instead translate 68k instructions and the rest of the CPU would have remained the same.
I think Motorola put their money behind PowerPC, a RISC architecture. I think it was a smart move on their part, because, while PPC didn't do so well on the desktop, it got into embedded products like video games and network hardware.
Today, the main CISC competitors to Intel are AMD and VIA, both who clone the Intel.
For a long time, the Motorola 68K family was ahead of Intel in every respect. What it didn't have was the whole PC industry depending on it and feeding it with endless financial resources.
The Motorola 68k line really did reach the end of its rope in the 1990s and everybody from Sun to Apple to Amiga had to plan an exit strategy or go out of business.
The Intel x86, on the other hand, continues to outpace everything else on practical performance. ARM is coming on strong in tablets and smartphones, but overall, the x86 platform has lasted a long time.
Not long after he wrote this, IBM's System/360 architecture went through an amazing transformation that's kept it competitive.
From the beginning, the System/360 was based on bipolar logic that consumes much more power than the CMOS logic microprocessors are based on. In the early 90's, IBM realized that further development of bipolar logic was impossible because of power and heat dissipation issues.
IBM had to go to CMOS, but CMOS logic wasn't as fast as bipolar logic at the time, leaving IBM unable to produce mainframes as powerful as the last generation -- unable to handle the requirements of existing customers.
IBM's answer was "Parallel Sysplex", a clustering solution that presents a single-system image of a group of mainframes. With Parallel Sysplex, IBM produced microprocessor-based mainframe clusters that scaled beyond the last generation of bipolar mainframes.
Some architectures end up being dead ends, but if something is commercially successful, it's amazing how long it can sometimes be strung along.