Ex Tenebris Lux & Revealing Everspin

There’s a Latin phrase I like; Ex Tenebris Lux—from darkness, light. Well, there’s lot’s of activity buzzing here in spite of the economic malaise. Magen was acquired by PPD. Our newest portfolio company Luxtera, the leader in CMOS photonics (optics on chips) announced a deal with Freescale, from whom we also spun-out a breakthrough company in nanoscale memory, Everspin, which is revealed below.

Right now: there’s an alphabet soup of economic trajectories: V, W, L, square roots signs. The slants and their slopes are shaped by investor expectations. Of course fundamentals matter—just less than expectations of fundamentals. Like bending a spoon with your mind—and often as illusory—the collective psychology of markets bends the shape of expected future outcomes. Exuberance and optimism inflect expectations upward. Despondence and pessimism yank them downward. A sunny day with expansive landscapes and vista views makes investors happy, calm and confident. It’s like standing upon a ridge, clear skies as far as the eye can see, investors lengthen their horizons, lower their discount rates and take long-term view on assets. Exploration and risk-taking are the order of the day—life is good. But when an unexpected storm full of fury hails down, so does panic. Fueled by fear, investors run for cover turning sad, nervous and uncertain. Huddled in the herd, staring downward barely focused beyond their footing, time-horizons shorten, discount rates skyrocket. Forget the future, the here-and-now triumphs, survival is the order of the day.

But here we celebrate those inventing the future. And today I share an exclusive Forbes sit-down with CEO of one of my venture firm Lux Capital’s most exciting portfolio companies, Everspin.

Aurangzeb Khan is president and CEO of Freescale Semiconductor-spinoff Everspin Technologies [full disclosure: my venture firm Lux Capital is an equity investor]. Earlier, he co-founded Altius Solutions, Inc., where he served as president and CEO through its merger with Simplex Solutions, Inc. and contributing to a successful initial public offering (IPO) in May 2001. He then served as an executive vice president and general manager at Simplex through its acquisition by Cadence Design Systems [CDNS] in June 2002. Aurangzeb has also held several engineering and general management positions at Cirrus Logic [CRUS], Tandem Computers (now part of HP [HPQ]) and Fairchild [FCS]. He helped deliver several industry-first systems and SoCs to market, several of which received industry-leadership business and technical recognition and have achieved $200 million to more than $1 billion in annual revenues. Aurangzeb received a master's in electrical engineering and a master's degree in engineering management from Stanford University, and a double-major bachelor of science in electrical engineering and computer sciences and nuclear engineering from U.C. Berkeley. Aurangzeb holds eight patents in IC circuit design, has contributed to more than 90 technical and business papers and has received coverage in more than 100 published articles.

Let’s start with Everspin. You are a Silicon Valley veteran, what attracted you to join as CEO of the company?
Everspin pioneered the integration of nanomagnetics with semiconductor devices. Many companies have tried to build this kind of memory (MRAM), but the folks at Everspin were the first to pull it off. I had been following the team’s work for a while, and what I particularly liked about Everspin’s approach, besides the leading-edge innovation, was that they did everything with commercialization in mind—with a view of producing something that could be manufactured in high volume with high yields. So that was a big part of what attracted me to the company.

Also, over the last few years I’ve done a lot of work in system-on-chip design, including work with Sony [SNE] for the PS2, as well as other consumer, networking, and telecommunications devices. I realized that in all of these devices, memory plays a critical role in the performance of the systems, so creating a breakthrough new memory could make a dramatic impact. I saw Everspin facing an enormous market opportunity, and that excited me.

By joining the company, what were you hoping to do?
I wanted to help the company grow and scale, bringing a bit of my own experience in both founding and building new companies as well as contributing at large organizations that brought new, industry-leading products to the market.

Let’s back up for the readers - define MRAM and why it’s important.
MRAM is a type of computer memory that consists of a nanomagnetic tunnel junction device and a single transistor to read and record data. It’s similar to other types of memory, except for the key difference that it’s non-volatile. That is, if you take the power away from this device, it doesn’t lose its memory, and that’s because the information is stored as a magnetic state, not as an electric charge. Physically, it is a very simple design, so there is a lot of headroom to scale the technology forwards.

What kinds of products do you envision MRAM being used for?
From the baseline technology we are building, there are three classes of products that emerge. The first are standalone memory chips. These are chips that you or I could go buy to install into our systems. The second is memory for system-on-chip applications, where Everspin’s memory could be integrated onto one piece of silicon with the other components necessary for a given system. The third application is sensors, and it comes from the fact that our devices are actually very sensitive magnetic sensors, relevant for consumer, automotive and other applications where high sensitivity is important. So those are the three markets we are able to address.

What are some of the early applications where Everspin MRAM is being used today?
In one application, Siemens’ [SI] Industry Automation division—a customer of ours for two years—is placing our chips into human-machine interfaces and programmable logic controls for industrial automation machines. In these types of applications, non-volatility is important. In the past, people would try to make a form of non-volatile memory by connecting a static, volatile memory chip to a battery. That way, if the power went away, the battery would have enough charge to write the data from the memory to some other more permanent location. We have heard from our customers that these types of systems are unreliable and often malfunction, whereas our memory is intrinsically non-volatile. Just recently, we issued a press release announcing that Siemens has shipped over 100,000 systems with our memory into the field without a single failure. They assessed our chips as having perfect quality and reliability, an accomplishment we are very happy about.

Another customer of ours is Emerson [EMR], a global leader in electric power support. Yet another customer uses our chips in casino gaming systems to keep track of user data, configuration settings, and transaction data.
We are also getting a number of design wins in avionics, and in one example the Japanese space agency is using an MRAM chip in a satellite launched into orbit early this year to measure carbon dioxide levels in the atmosphere.

When you think about the future of MRAM, what do you think of as the greatest market opportunity?
I think storage is a very good opportunity for us. In areas where mission-critical information needs to be kept available and reliably stored with integrity, our solution is excellent. From the consumer perspective, imagine being able to turn a computer or device off and on instantly! MRAM gives you that ability because when you power the device off, it retains all of the information, even if the battery life runs out.

I think instant-on computers would be an instant hit with the public. How far away are we from seeing that kind of capability?
Well our density today is probably inadequate to serve as the boot up for a full operating system, but if you look at our roadmap, it’s conceivable that in the next several years we will have a meaningful level of density where at least part of that function can be implemented.
Today, for example, people are already using our memory in devices like studio cameras to maintain configuration settings for fast content switching.

Everspin itself was a spinout of Freescale Semiconductor. Tell me about the relationship with Freescale and your plans to speed up MRAM adoption as an independent company now.
Freescale is an investor in Everspin - it has a stake in the venture and believes in the technology, but we are an independent company managing our own destiny. Today, we are essentially a tenant of Freescale, so if you came to visit, you’d find me sitting in a facility that is part of the Freescale campus. Everspin has its own nanomagnetic fabrication facility, and that fab resides in a clean room space built up by Freescale.

After spinning out from Freescale, a lot of our work over the last year has been focused on the sales and marketing aspects of our business. We now have a global network of representatives, distributors and direct sales staff to uncover and serve the new market opportunities that I mentioned. Once folks become aware of where MRAM technology is today and how far it has come into its production history, a lot of opportunities open up because it gives designers a flexibility that they didn’t have before. Our team has also been very busy working on the technology and product development aspects of the business, and now we have 50 different part numbers in our portfolio of products. We have continued to expand the product, with new kinds of interfaces and new kinds of density footprints. Later this year we will introduce both a serial peripheral interface and a 16Mb design that will further expand our lead in this market.

You are approaching the one-year mark of Everspin being an independent company. What would you say is the biggest accomplishment to-date?
Considering that we are going through one of the most severe economic challenges in the last 25 years, I am just delighted with the level of market and customer traction that we are seeing. I think that in itself is a huge accomplishment, and the Siemens zero-failure news that I shared earlier underscores an incredible amount of hard work by the whole team to reach this milestone of being a demonstrably reliable innovator and business partner.

What other things should people be looking for over the next year from Everspin?
We are actively developing another kind of MRAM technology that several other folks are working on, called spin torque (ST-MRAM). This team has been working on spin torque since the very early days of MRAM, and we will continue to develop that technology and integrate it into our roadmap at the right time.

Geographically, are you seeing increasing patterns of demand or competition in certain parts of the world that have surprised you?

We’re not seeing much competition in the MRAM space, because there is no other supplier with comparable capability in the market today. In terms of demand, we are seeing good growth in Europe and Asia and we are in the early stages of getting wins in Japan at a number of accounts. We are seeing a lot of interest from the major multinationals who will try us on a first product, and after they have had some experience with us and see how well the product works, we are seeing customers expand the use of our chips into other products in the same business units, and subsequently giving us introductions to other business units within those organizations.

When you think a few years hence, what do you hope everybody will look at Everspin and say?
Our goal is to bring breakthrough products to the market in a way that shows that we are not only a leading innovator, but a reliable business partner as well. I think those two aspects together are very powerful - the combination of bringing truly industry-first technology to market and doing in it in such a way that companies can depend on us to be there with the right product when they need it is pretty special. That’s a big part of the Everspin brand: agile, responsive, and customer focused so as they are successful, we have a chance to succeed as well.

Who is your favorite author of all time?
Growing up I read a lot of Bertrand Russell. I like the fact that he was very inquisitive and also an iconoclast – he didn’t just take the conventional wisdom as received knowledge, and I guess I aspire to be a little bit like that.