Intel predicts Moore's Law to last another 10 years

Intel is gearing up to celebrate the 50th anniversary of Moore's Law this month, and has kicked off proceedings by claiming that many of the achievements of the computer age can be attributed to the rule, and that the firm still sees no end in sight even as it looks beyond 10nm chips.

The anniversary of the first publication of Moore's Law falls on 19 April, and Intel held a briefing at which Intel senior fellow Mark Bohr said he didn't see Moore's Law slowing down in the near future, even as the firm develops a 10nm production process and is researching 7nm and 5nm technologies.

"We can see about 10 years ahead, so our research group has identified some promising options [for 7nm and 5nm] not yet fully developed, but we think we can continue Moore's Law for at least another 10 years," Bohr said.

However, he added: "If you had asked me 10 or 20 years ago, I would have given you the same answer: our visibility back then was about 10 years into the future. It's fair to say that eventually Moore's Law will slow down or come to an end, but we don't see that in the imminent future."

Bohr predicted that Moore's Law will not come to an abrupt halt, but will morph and evolve and go in a different direction, such as scaling density by the 3D stacking of components rather than continuing to reduce transistor size.

For example, Intel announced 3D Nand flash memory technology last month jointly developed with Micron that is claimed to offer three times the capacity of existing components.

In response to a question on whether optical or photonic computing might replace silicon-based processors, Bohr expressed scepticism that this would happen anytime in the foreseeable future.

"I'm not very optimistic about photons or optical interconnects being useful for on-chip computing, simply because the energies and the wavelengths used are not very competitive with electrons and normal copper wires," he explained.

"I think electronics will stay the mainstream for the foreseeable future on chips, while optical interconnects as a means to communicate chip to chip is going to become increasingly important as we go forward."

Bohr was cagey about the materials Intel is looking to as it moves to ever smaller components, but said that the basic transistor has already changed from silicon oxide to one that uses a hafnium-based dielectric, and that Intel is looking at so-called 'III-V' materials from the third and fifth columns of the periodic table for possible use in the transistor channel.

Intel innovation strategist Steve Brown tried to claim that Moore's Law was responsible for most of the advances in the digital age, from personal computers to supercomputers used for medical research to discover new treatments.

"If you look at the very first Intel microprocessor, the 4004 from 1971, today's transistors have 3,500 times the performance, run 90,000 times more efficiently and we are able to make them at 60,000 times lower cost," he said.

"If none of that had happened, an Intel-based Android phone would have a processor the size of a parking space."

Bohr put some perspective on that claim, explaining that the existence of Moore's Law, which posits that the number of transistors that can be crammed onto a chip will double every two years, gave Intel engineers a target to aim for.

"If we didn't have that articulation of a regular cadence of reducing feature size, the whole industry would be much further behind," he said.

"It's hard to speculate what that would be, but I think it is fair to say that we wouldn't be where we are today with 14nm dimensions and chips with over a billion transistors."

• Tweet  
• Facebook  
•  
•  
• Save this article  
• Send to  

• Topics
• Processors
• Datacentre
• Components
• Intel
• Chips