A years-long mission to build a microprocessor out of carbon nanotube transistors has finally succeeded thanks to a team of MIT researchers.
The development comes as the sustainability of Moore’s Law is increasingly called into question. Silicon-based transistors are nearing the point when they will be unable to shrink anymore, delivering increasingly marginal improvements.
As a result, a carbon nanotube-based microprocessor has been a design goal for next-generation computing for some time. Researchers believe that these nanotube transistor-based designs will deliver far faster performance with around 10 times greater energy efficiency – and as a result, fewer environmental impacts than silicon-based microprocessors.
The carbon nanotube field-effect transistors (CNFETs) at the heart of this new era of microprocessors have one critical flaw: when produced at-scale, they have defects that render them less performant and unsuitable for mass consumption. Most notably, a small portion of fabricated nanotubes will be metallic, interfering with how the transistor switches.
This is what MIT’s breakthrough addressed. Their new technique – called “DREAM” (“designing resiliency against metallic CNTs”) – positions those metallic CNFETs in a manner where they aren’t disruptive to the transistor. This reduced the purity requirement from 99.999999% purity to 99.99% purity – four orders of magnitude.
“The ‘DREAM’ pun is very much intended, because it’s the dream solution,” said co-author Max M. Shulaker, assistant professor of Electrical Engineering and Computer Science (EECS) at MIT. “This allows us to buy carbon nanotubes off the shelf, drop them onto a wafer, and just build our circuit like normal, without doing anything else special.”
With their DREAM technique, the team was able to demonstrate a 16-bit microprocessor with more than 14,000 CNFETs (pictured in the header image) that successfully performed the same tasks as silicon-based microprocessors. The researchers had their new microprocessor run a variation on the iconic “Hello, World!” program. (“Hello, World! I am RV16XNano, made from CNTs.”)
This microprocessor iteration, which is based on the RISC-V open-source chip architecture, builds on a previous version designed by many of the same researchers six years ago – but that iteration only had 178 CNFETs.
“This is by far the most advanced chip made from any emerging nanotechnology that is promising for high-performance and energy-efficient computing,” said Shulaker. “There are limits to silicon. If we want to continue to have gains in computing, carbon nanotubes represent one of the most promising ways to overcome those limits. [The paper] completely re-invents how we build chips with carbon nanotubes.”
Now, the researchers are setting their sights on real-world applications, beginning with the implementation of the DREAM technique in a chip foundry through a Defense Advanced Research Projects Agency (DARPA) program. But with the new technique, the future is bright for a new generation of computing.
“We think it’s no longer a question of if,” Shulaker said of commercially available carbon nanotube processors, “but when.”
Read the MIT announcement of this news here.