The October 2022 issue of IEEE Spectrum is here!

Close bar

Carbon Nanotube Memory Company's Ship May Finally Come In

An analyst's report says Nantero appears poised to cash in on its quixotic journey to bring carbon nanotube-based nonvolatile memory to market

5 min read
Nantero NRAM nonvolatile memory chip
Illustration: Nantero

To many, the sell-by date on the carbon nanotube-based non-volatile random access memory (NRAM) developed by Nanterohas long since passed. IEEE Spectrumcharacterized the technology as a “loser” nearly a decade ago after several of the company’s launch dates came and went with hardly a whimper.

The technology’s promise was that it could lend itself to easy mass production because it relies on a group of nanotubes deposited randomly on a substrate rather than individual nanotubes precisely placed. By eliminating the need for individual placement, Nantero hoped to sidestep the main bugbear of nanotubes in electronics: purity. It turns out purity could not be sidestepped to the degree originally believed, and a decade-and-half of disappointment ensued.

But a new analyst report published by BCC Research asserts that NRAM’s ship may have finally come in. The upshot: It (and a host of other non-volatile memory approaches) may be poised to dislodge flash from its long-held throne.

The long and torturous journey taken by Nantero in bringing NRAM to this point  is a story of setbacks, perseverance, and what appears now to be success, if BCC Research’s analysis is on the mark. 

“The long runway for NRAM and Nantero can be seen as a feature and not a bug”

The brief description of this odyssey includes US $78 million in financing over 15 years and $50 million in revenues over that period through licensing of its technology and patents. That revenue includes a back-against-the-wall sell-off of a key business unit to Lockheed Martin in 2008.  Another key moment occured back in 2012 when Belgian-based nanoelectronics powerhouse Imec took on the job of further developing Nantero’s carbon-nanotube-based memory back in 2012. Despite the money and support from major electronics players, the big commercial breakout of their NRAM technology seemed ever less likely to happen with the passage of time.

So, what triggerered the sudden turnaround in Nantero’s fortunes and caused analysts to take another look? According to BCC, it was primarily a licensing deal struck this past summer with Fujitsu.

Of course, this was not the first licensing deal that Nantero had entered into over the years, and licensing deals for disruptive technologies are always a shaky proposition. But Chris Spivey, senior editor at BCC and co-author of the report, told IEEE Spectrum in an e-mail interview that there was something unique about this deal.

“Fujitsu announced that they see NRAM as having a natural place alongside their ferroelectric random access memory (FRAM),” says Spivey.

“No one had processed carbon nanotubes to less than 1% of impurity, and we discovered we needed to be at less than 10 parts per billion ”

Spivey explains that he had been told by Takashi Eshita, senior director for system memory at Fujitsu, that although FRAM has excellent advantages, such as high speed, low power consumption, high endurance, and so on, shrinking its size is slightly difficult. NRAM has the potential to be ramped up to large memory capacities, so with that in mind, Fujitsu decided to boost its non-volatile memory lineup by starting to develop NRAM for large-scale memories.

While this latest deal with Fujitsu is what turned people’s heads, Nantero may have begun to right the ship back when it sold off its key business units to Lockheed Martin, according to Spivey.

“The key historic licensing deal was not for computer memory but for advanced materials, which is really how I now categorize Nantero—a materials company whose first major public products happen to be in computer memory applications,” says Spivey. “That key license, back in August 2008, gave Lockheed Martin exclusive license arrangement with Nantero for government applications of the company’s intellectual property portfolio. I believe the sacrifice of chopping off a limb was painful for Nantero, especially offering up employees, but they likely had zero alternatives back then.”

It wasn’t all licensing deals and strategic business decisions that may have turned around Nantero’s fate. The BCC report does make mention of some technology tweaks that have helped NRAM get over the hump.

While not giving much in terms of specifics, the report explains that Imec’s contributions to NRAM starting back in 2012 helped to shrink the memory cell size and in so doing offer better scalability to sub-20 nanometer nodes.

This essentially created two generations of NRAM with the first being a three-terminal semiconductor device in which the third terminal was used to switch the memory cells between various memory states. The second generation of NRAM technology was based on a two-terminal memory cell and enabled the smaller size.

Undoubtedly, the biggest technological hurdle for NRAM and the greatest accomplishment for Nantero was overcoming the purity issue of the carbon nanotubes. The BCC report offers some interesting quotes from Nantero CEO Greg Schmergel that highlights what a herculean task the purity issue ultimately turned out to be when at the beginning it was thought to be already solved.

“No one had processed carbon nanotubes to less than one percent of impurity, and we discovered we needed to be at less than 10 parts per billion impurities. So that’s a big difference,” Schmergel is quoted as saying in the BCC report.  “So we needed to invent that process, and now we’ve got it down to less than one part per billion…and that’s well protected by patents. More than 10 years later we are still the only company that can take carbon nanotubes into a production CMOS fab without risk of contamination. So that’s given us a huge advantage for many years [to come].”

As the BCC report explains, the purity issue was an obstacle that if not overcome could have buried the prospects of NRAM. However, an overarching issue in the technology’s development was the demand from industry that the technology must never require any special processing outside of standard manufacturing.

“The marching orders from the fab were: we won’t buy any new equipment for you, we won’t modify any equipment for you,” Schmergel.  “We won’t move anything for you. We won’t learn a new process for you—we won’t do anything. So you must fit your manufacturing process 100% into what we do.”

Needless to say this uncompromising position took a great deal of time to satisfy. Nonetheless, paradoxically, it may been the long delays brought on by these issues that ultimately benefited Nantero in fighting the Goliath of flash memory. Fifteen years ago when NRAM was being promoted as the next big thing, flash memory still had a lot of room for improvement. That situation couldn’t be more different today, when many will tell you that flash has no more legs left and the time has come for new non-volatile memory solutions.

“The long runway for NRAM and Nantero can be seen as a feature and not a bug,” says Spivey.

With the passage of time, according to Spivey, it was possible to look at NRAM with a fresh set of eyes that highligthed its technical strengths rather than merely from a perspective of commercial disappointment.

Spivey adds: “CNT’s have inherent advantages as memory media—they can be packed in high density, they operate at extremely high temperatures, and most importantly, at picosecond speeds. And they comparatively require little operating energy.”

Technology tweaks, key business decisions and the passage of time all appear to have set the stage for NRAM to finally have its day.

The Conversation (0)

3D-Stacked CMOS Takes Moore’s Law to New Heights

When transistors can’t get any smaller, the only direction is up

10 min read
An image of stacked squares with yellow flat bars through them.
Emily Cooper
Green

Perhaps the most far-reaching technological achievement over the last 50 years has been the steady march toward ever smaller transistors, fitting them more tightly together, and reducing their power consumption. And yet, ever since the two of us started our careers at Intel more than 20 years ago, we’ve been hearing the alarms that the descent into the infinitesimal was about to end. Yet year after year, brilliant new innovations continue to propel the semiconductor industry further.

Along this journey, we engineers had to change the transistor’s architecture as we continued to scale down area and power consumption while boosting performance. The “planar” transistor designs that took us through the last half of the 20th century gave way to 3D fin-shaped devices by the first half of the 2010s. Now, these too have an end date in sight, with a new gate-all-around (GAA) structure rolling into production soon. But we have to look even further ahead because our ability to scale down even this new transistor architecture, which we call RibbonFET, has its limits.

Keep Reading ↓Show less
{"imageShortcodeIds":[]}