Of course PC Microprocessors are faster every year Mike but that’s not the point. There has been competition in the PC microprocessor market all along. And that is exactly my point. If Intel had exclusive rights to microprocessors used in PCs, do you think they would be as good today as they are without any competition to spur them to advance? I believe they would not. People demand faster and cheaper but if only one company is allowed to control the hardware architecture, just how much innovation will happen? There’s not much to drive them to innovate. Drive cost out yes, but innovation? Why spend R&D dollars for that when you have a captive market? To give one company exclusive rights to hardware designs used in an emerging technology is casting the mold to enrichen just one company over all others. It’s not good for innovation and it’s not fair to everyone else trying to get a piece of the pie.
There’s direct evidence of that. Back in the day, Intel was the only game in town unless you wanted a Macintosh, which used Motorola chips. They were kinda slacking for awhile until AMD got serious about competing with them. Now Intel has some truly excellent products (which keeps AMD on its toes), and Apple switched to them instead of Motorola to boot.
The competition angle is definitely important. There’s an book people here might be interested in reading: “The Yugo / The Rise and Fall of the Worst Car in History.” There’s a section that talks about Yugo-America going over to the main Yugo factory (Zastava Automobiles) and being astonished at how behind the times it was. And not just out-dated, but sloppy and lazy to boot.
But Yugo had never had to compete. The government told them to build cars, and then gave them money whether they built them right or not. There was no incentive to get better.
Once the Americans explained that such things wouldn’t fly over here, they stepped things up in a big way. The Yugo we got over here was orders of magnitude better than what the factory had been cranking out previously (just think about that for a moment! 
) They still couldn’t hack it because you don’t go from Yugo to Honda in 6 months, but it’s an interesting insight into how competition forces every product to get better or die.
The book’s also kind of fun to read because it goes in-depth about Malcolm Bricklin, the PT Barnum of the car world. He’s the guy that brought Yugo here. Same guy who brought Subaru here, to considerably better success (probably because it very quickly went public and so Bricklin couldn’t wreck it).
Other than the Subaru venture, pretty much everything he touched died on the vine, including his godawful Bricklin Safety Car.
Intel and AMD keep leap frogging each other. But they still run the same instruction set. Different implementations of the same instruction set so they can run the same programs.
Moore’s law isn’t dead yet. It may have slowed down for a little while. Doesn’t mean it’s dead. Many are still predicting it could start to speed up again.
Yes…because people needed it and demanded it. Software has been the big driving factor for chip manufacturers making faster systems. Consumers want faster systems. If Intel didn’t make them then the market wouldn’t grow.
I was moving from mainframe and super mini programming to programming on PC’s during that period…and Intel’s chips were easily doubling in speed every 18 months. Intel’s big leap came when they stole Digital Equipment Company’s Alpha chip design. DEC took them to court and won. And instead of Intel paying DEC - they bought the technology. At the time DEC made the fastest chips. Their floating point processor was well over 1,000 times faster then either Intel or AMD.
There are few points being completely missed here.
. I never said competition is bad.
. You can still have competition building systems with the same instruction sets for standardization.
Microsoft (or anyone else) doesn’t write two different pieces of software for Intel and AMD. It’s exactly the same. Only difference would be for the 32 bit (pretty much dead) or the 64 bit. Windows and Linux (and all underlying software) are exactly the same.
So we have a standard (x86 instruction set) and competition by two chip manufacturers with different architecture approaches.
The PC market has standardized around the X86 chips…But there is competition in that space.
I want to share your optimism, Mike, but the realist sitting inside does not see any indication toward this yet.
I do agree that in general performance will improve, but not from the Moore “law”, which is more of statistical observation, than something rooted in reality.
Here is another one of the kind:
We may be close to Moore’s law in current Silicon chip technology. But there are other technologies that are coming of age to take us to next level.
Compound Semiconductors for one. Very promising. Graphene is also very promising.
The last time I will point out the obvious. It’s right in the linked article. Competition will be stifled:
The AVCC’s recommendations will likely include products made by some or all of its members. Among other things, it could turn out to be a way for companies like NVIDIA and NXP to own a large share of the future market for self-driving-vehicle system components – and, potentially, to shut out smaller or less-nimble rivals.
Last I knew…I’m not the one who wrote the article. I stated my opinion on my 45+ years experience in this field.
No need to state the obvious.