HP reckons that with developments going into chip functionality at an atomic level, reaching the end of Moore’s Law is “irrelevant”.
Discussing the future of nanoscale computing at a Kavli Foundation roundtable, HP Research Fellow Stan Williams, said that the importance lies in how information is stored and moved.
“People talk about reaching the end of Moore’s Law, but really, it’s irrelevant,” Williams said.
According to Williams, transistors are not a “rate-limiting factor” for today’s computers, and that improving transistors by “a factor of one thousand” would have no substantial impact on the modern computer.
“We can continue to improve data centres and computers at Moore’s Law rates – doubling performance every 18 months – for at least another 20 years without getting into something like quantum or neuronal computing,” Williams said.
While it is fun to think about computers that compute more like a brain, he says, it will be “20 years before any of that is actually needed” as there are so many other improvements that are needed first.
HP also expects to have commercially available memristor based flash by the end of the year, though it could cannibalise its development partner Hynix’s own flash products.
According to Williams, HP’s drive to create a viable product based on memristor technology is almost coming to fruition.
Memristors, which can ‘remember’ previous current levels, have been theorised since 1971, with HP leading development of the technology.
Speaking as part of a Kavli Foundation roundtable discussion, Williams said the firm is nearing completion of memristor based flash devices.
“In terms of commercialization, we’ll have something technologically viable by the end of next year,” Williams said, pointing out that the economics, investment, and market readiness are more difficult to figure out than the science itself.
He says that this could cause some problems however for its development partner Hynix, meaning that timing is important from a supply and production point of view.
“Our partner, Hynix, is a major producer of flash memory, and memristors will cannibalise its existing business by replacing some flash memory with a different technology,” he said.
This means that the way the firm times the introduction of memristors turns out to be important, he says.
Williams points out that there is a lot more cash thrown at shaping the market than is ever spent on developing the technology in labs.
“Development costs at least 10 times as much as research, and commercialization costs 10 times as much as development,” he said.
He added: “So in the end, research — which we think is the most important part — is only 1 percent of the effort.”
He also pointed out that HP will not be making money directly from memristors, but will seek to “ensure there is a supply chain”.
“We’re not going to make money off these chips. We are going to make money by building cool systems utilizing these chips.”
“Hopefully, we’ll build cooler systems than other people because we’ve been thinking about this technology longer.”