TORONTO — Better late than never might be a good way to sum up NRAM.
After years of not quite be ready for wide adoption, a new report from BCC Research is predicting that Nano-Ram (NRAM) is finally in a position to disrupt incumbent DRAM and flash memory with commercialization expected in 2018. The Wellesley, MA.-based research firm said the first non-volatile memory chip to exploit carbon-nanotube technology looks like it's finally ready to have a serious impact on computer memory.
“Industry experts had given up on waiting for CNT memory," said BCC Research editorial director Kevin Fitzgerald in an interview with EE Times. “I believe one needed fresh eyes to really see that the time was coming when it was really possible to make the switch from silicon to carbon."
He said a recent licensing agreement to manufacture NRAM commercially by tech giant Fujitsu prompted the firm's change in outlook, and while the report was being written, the NRAM vendor that struck that agreement, Nantero, received an additional $21 million in venture funding, solidifying the firm's suspicion NRAM is about to break into the open. BCC Research anticipates the global NRAM market to achieve a compound annual growth rate (CAGR) of 62.5% from 2018 to 2023, with the embedded systems market anticipated to reach $4.7 million in 2018 and $217.6 million in 2023, growing at a CAGR of 115.3%.
The independent report expects NRAM technology to be more disruptive than flash memory was to enterprise storage, enterprise servers and consumer electronics, enabling a wave of innovation in those products. Applications that will be impacted quickly will include the consumer electronics sector, mobile computing, Internet of Things, enterprise storage, defense, aeronautics and automotive.
Fitzgerald said several computer memory experts and interested observers were already bullish about non-memory uses for the carbon nanotubes. “Many of the applications are directly related, like chemical sensors or RFID repeaters, but several are expressly for advanced materials uses, such as solar cells, fuel cells/batteries, power transmission and MEMs in general," he said “The deeply stringent semiconductor manufacturing requirements has, it seems, led to unanticipated advantages for other applications. And from a business perspective, highly valuable IP positions."
NRAM has a lot of factors working in its favor, said Fitzgerald, including greater density. “It requires far less energy as it does not need to constant 'refresh' that traditional memory does. It can be made fast, potentially down to the picosecond response time." It also allows a true “instant on" capability, and is fantastically heat tolerant, he added. “It really should not wear out in the traditional sense."
Fitzgerald believes Nantero's decision to license technology rather than go it alone was a key tipping point for pushing NRAM forward. “This enabled Fujitsu to step in and move the manufacturing at an accelerated rate," he said. “With a larger part of the pie available, partners are motivated to take on a share of the product development burden."
From a technology perspective, purity of carbon source was a major hurdle, said Fitzgerald. “Solving that put Nantero in a strong IP position, and it can been seen as a material platform company whose initial products happen to be in the computer memory sphere."
Nantero CEO Greg Schmergel also highlighted the Fujitsu agreement and its recent funding as key milestones for the company. He said the biggest opportunities and areas where the company has active partnerships is in the mobile and consumer flash memory application space, and expects substantial business in mobile and tablets. “We already have customers there. The value proposition is strong for these devices because battery life is key."
Schmergel said non-volatile memory with the speed of DRAM means far less power consumption for devices, and users will notice the speed improvements. “You are no longer bound by typical tradeoffs."
He said some of Nantero's partners are designing NRAM products, while the company itself is doing some of its own design, including a DDR4 compatible memory product. It now employs more than 70 people.
BCC Research has heard of several standalone memory companies, phone and wearable device companies that may get involved with NRAM, but Fitzgerald said they remain anonymous for now.
The path of NRAM is rather unique from any other memory technology, said Fitzgerald, even those that have taken time to gain traction commercially, such as MRAM. “None of the traditional memories have had this kind of back story. Moving from silicon to carbon meant carrying a lot of extra baggage along with it, but several companies are exploring CNT memory, IBM being one of the larger entities."
Nantero's proprietary NRAM is based on carbon nanotubes
Chris Spivey, BCC Research co-founder and senior editor who co-authored the report describes the research and development path as a classic “David versus Goliath" story, but with a twist, in that ultimately, Nantero as David opted to enlist the help of one of Goliath's cousins, Fujitsu. Meanwhile, he said the silence that is coming from incumbent memory companies when approached during the writing of the report is telling, and suggests to him they may be in fact going through the motions of licensing.
As noted in the BCC Research report, the obituary for NRAM has been written and published half a dozen times, making for such headlines as “NRAM passes its sell by date" or simply “Loser." However, the authors anticipate this significant technology advance to become a direct and successful challenger to entrenched computer memory technologies. “Part of our thinking (sticking to biblical metaphor) is that the end of the dominance of flash memory has been written on the wall by the moving hand for several years, and the ink is now quite dry. And the moving hand has also begun scribbling unhappy thoughts about DRAM."
The report also quotes Symbolic IO senior VP and CTO and former long-serving Micron exec. “I believe the opportunity is very large for several new types of non-volatile memory to emerge beyond 3D Xpoint – NRAM being one of the obvious – but there are several hopeful entrants. I will be delighted if any or all of them are successful."
BCC Research's Fitzgerald said the memory community has been waiting for 25 years for something better to arrive. “In some ways NRAM is in the 'better late than never' category. This is an industry thirsting for change."
—Gary Hilson is a general contributing editor with a focus on memory and flash technologies for EE Times.