Intel says chips to become slower but more energy efficient

[Sabre]

The Stack

Intel has said that new technologies in chip manufacturing will favour better energy consumption over faster execution times – effectively calling an end to ‘Moore’s Law’, which successfully predicted the doubling of density in integrated circuits, and therefore speed, every two years.

It’s a prediction worth remembering, since Gordon E. Moore himself was the co-founder of Intel and Fairchild Semiconductor when he made the prediction that led to ‘Moore’s Law’ in a paper [PDF] back in 1965.

The prognosis comes from William Holt, Intel’s Executive Vice President and General Manager of its Technology and Manufacturing Group, speaking at the International Solid State Circuits Conference in San Francisco, and discussing the new technologies – such as tunnelling transistors (or ‘Quantum tunnelling’) and spintronics – which will define the next stages of evolution in computing.

“We’re going to see major transitions,” said Holt. “The new technology will be fundamentally different.” and continued “The best pure technology improvements we can make will bring improvements in power consumption but will reduce speed.”

Holt elaborated that while Intel recognises the need to consider re-tooling its plants and committing to new technologies in chip production, it hasn’t made a decision about direction yet. Quantum tunnelling, though brought to advanced proof-of-concept by DARPA and the Semiconductor Research Corporation, is currently further from commercialization than spintronics, which uses quantum mechanical properties of particles as switch facilitators, and which is expected to begin to appear in commercial technology such as graphic chips within 18 months.

Well, the law had to be broken some time. I really think that the lack of good parallel computing software (IDE's, compilers, etc.) will continue to hinder us for awhile. It's an egg that no one has really been able to crack.

[drwrx]

The big problem I saw some time back was that there really isn't a "need" for such processing power outside of a few industries. For example, my work with huge multi gigabyte photoshop files is not really hampered by processor speed any longer. Yes, I see speed improvements. Going from 10 seconds to 5 seconds applying filter applications on huge files is nice, but honestly at this point, I'm the bottleneck. Opening, saving and closing are still the big slow downs and those are not really processor issues. SSDs have made the biggest difference in those cases. But ultimately, you need a certain level of demand to drive progress and it's just not there.

Do we think that the general public needs a 12 core 4 GHz computer with 64 GB of RAM to run Excel, Word, PowerPoint, Outlook and Chrome? Sure, we would all love it, but are we willing to pay $9k for it? That is the kind of cash that pays for pushing the processor envelope and there isn't any need for it at a large scale.

[Sabre]

Oh, I agree. About the only thing that "normal" people do that requires a good bit of computation is play games, and 99% of that is handled by the GPU now'a'days (which consequently, are huge parallel processors and have advanced parallel computing more than anything).

SSD's have changed how fast computers feel, that's for sure. Disk has always been the slowest thing.

The next big things will be HBMincreasing memory speeds and increasing CPU caches to huge amounts and XPointincreasing the speeds of SSD by orders of magnitude. Machines will feel MUCH faster with these technologies without ever touching the CPU.