Bosch Sensortec, InvenSense and STMicroelectronics share more than 70% of the inertial sensor market for consumer applications (more detail in the recent report Sensors for Cellphones and Tablets from Yole Développement) – but in the Chinese market a strong number four has emerged: mCube. The San Jose, California, headquartered company offers the smallest, lowest power consumption, 3-axis accelerometer on the market. Along with interesting new features, this is helping mCube target Internet of Things (IoT) applications. We have interviewed Mr. Ben Lee, chief executive officer of mCube, in order to better understand the company’s strategy and its future evolution.
Yole Développement (YD): I understand 2015 was a good year for you in terms of growth, financial gain, and moving to higher volume production.
Ben Lee: Yes, we are in this growth phase. The momentum seems to be building, particularly in China. It’s one of those things where the more people use mCube, the more people want to use mCube.
YD: mCube’s business since 3 years seems to expand exponentially, can you share the big picture of mCube?
Ben Lee: mCube’s strategy was to initially focus on the China accelerometer market because it was the fastest growth and potentially largest TAM. That proved to be the correct decision. The company has managed to double shipments from Q1 2015 to Q1 2016, after doubling in the same period the year before. With this volume, mCube is able to establish reliable production capacity, drive down cost and improve quality. This positions the company to expand into larger customers and new market segments.
YD: Clearly, yes. Your primary product is 3-axis accelerometers?
Ben Lee: Right, we’re very focused on the accelerometer, for good reason. If we look at the total market in China for accelerometers, it’s about 800 million units/year. The growth has flattened out recently, but it’s a very large market, and we were a fairly new player with small market share. The way we look at it is, the upside is very big for us and there is quite a bit of headroom for us to grow. As a company, we decided that while everyone else is de-focusing on accelerometers, we increased the focus, and that strategy has worked out well. I think we need to do one thing well. MEMS, as you know, is difficult. It’s not only hard to do, but it’s very hard to produce in volume. Because we focused on the accelerometer and now we have it in solid play, we can start thinking about expanding our portfolio.
YD: So the idea is to be extremely focused on this one device; to really compete in terms of size and price, and then be able to expand into other fields when you have enough volume of sales for accelerometers?
Ben Lee: Yes. I see a lot of companies that are starting out – or even a similar size to us – trying to do too many things. I think it’s not too hard to do a proof-of-concept. We actually have a lot of different projects with proof-of-concept that we have not introduced. But then with MEMS, it’s so hard to get a product into volume production. Once you ship your first million units, then you realize what the problems are: it’s even harder to get it to ten million units and 100 million units. A lot of companies have talked about good technology, but very few have been able to produce in high volume. So that’s the lesson that we’ve learned and the strategy that we’ve taken: do one thing well first, and then move on to other things.
Different architectures of accelerometer using TSV
(Source: Sensors for Cellphones and Tablets 2016 report, Yole Développement, June 2016)
YD: We see the market gradually moving to 6-axis and 9-axis IMUs, but there is still a market for accelerometers. It’s a market that will stay stable, but at a high value, so it’s a focus that really makes sense. There are also a lot of IoT and wearable applications that require only a 3-axis accelerometer.
Ben Lee: It’s very interesting that you say that. Even though we doubled every year for the last two years, we still can easily double again and still have headroom to grow. It will take maybe another year and a half to do that, so we have a nice growth trajectory for a year and a half. Our original thought was we’re going to hurry up start the combo devices. But what was interesting is that when we introduced the low-power accelerometer late last year in Q4, we found all of these opportunities in IoT and wearables that don’t need a gyro, and what they care about is power and size. So that’s a new growth market for us, leveraging the same technology platform and our ability to extend it into a new market. That’s exciting for us. Ever since we introduced our new IoT product, we’ve had customers all over the world approaching us, cold-calling us, flying in to see us, telling us that they had to use MCube, because no other accelerometer in the market can meet their needs. That’s very interestingbecause our previous focus was very much on China, on the smartphone market, but with this IoT market, we’re being approached by all these customers from different market segments and also potential partners. So now we’re putting more energy in expanding in this IoT area, leveraging our technology.
YD: We’ve published a new report on the status of the MEMS industry, where we highlight that one way to recover from the quarter-after-quarter price decrease is to provide new functionality, such as for IoT applications. For this, you need a device that is small, low-cost and extremely low power consumption. The 6-axis and 9-axis products don’t make sense here. Do you agree?
Ben Lee: Yes, I think you’ve hit it right on the nail here. We cannot, as a small company, compete on price alone. There are two fundamental things that we are doing. One is to drive costs down with technology, which we have accomplished. The second complementary leg of that is to add value at a software system level to the sensor. A very challenging example is when our customers say: ‘Hey, my power budget for a pedometer has to be less than 5µA. By the way, I want 90-95% accuracy.’ It turns out that it’s very hard to balance the two. If you push accuracy you may sacrifice power, and if you reduce the power consumption you may not get the accuracy. A lot of people can write algorithms. The reason we’re able to do a better job is that first of all, we know our sensors very well, and what software/hardware methods we can use to reduce the power. We write algorithms to leverage that, such that we are now delivering value that nobody else can, with a combination of power and performance. That’s why, at CES, where nobody wanted to look at just a chip, we decided to do the whole activity band, and that afforded us tremendous learning. We found that to get the total system power down it’s really a combination of sensor, algorithm and system integration working altogether with the MCU. So now I think we know better than a lot of MCU companies what it takes for an MCU to be low power at a system level. We have customers come to us and say ‘Wow, you’ve solved 90% of our problem, why don’t we just license the whole solution from you.’ I think that’s the key: find ways with specific applications to add value so that you’re not giving the sensor away.
YD: What do you think about the status of your competition? InvenSense, STMicroelectronics (ST) and Bosch Sensortec (Bosch) are fiercely competing with each other and your company. Each of these companies has quite different momentum at the moment.
Ben Lee: It was quite surprising; we did two exhaustive bottom-up surveys, one in May of 2015 and the second one in March of 2016. We looked at the entire accelerometer market. In May of 2015, Bosch was number one and ST was number two. Kionix was number three and we were quite a distance away. This year, we knew that we had gained a lot of market share, but we didn’t realize that ST had also gained so much. ST has now become the clear number one in China. Bosch has lost a ton of market share, they’re still number two, but we are a close number three now. When we looked at the Chinese smartphone sensor market, only about 15% use a 6-axis gyro; the remaining 85% of the market does not use one. At the mid-to-high end, about 20% of the market uses a compass of some sort, most of which are discrete, not combos. We don’t have 6-axis gyro+accelerometer, we only have 3-axis accelerometers, so our serviceable addressable market is about 80% of the total addressable market. Currently, we don’t compete directly with Invensense since they don’t have a standalone 3-axis accelerometer.
YD: You are able to compete with big players like Bosch, STM, InvenSense, by focusing on consumption, size, packaging and offer great solutions, what are your key advantages, your strength?
Ben Lee: mCube’s key advantage is our technology leadership. With our monolithic single-die platform, we continue to challenge the laws of physics and stretch the limits of miniaturization. Our latest example is the award winning world’s smallest MEMS 3-axis accelerometer in less than a cubic millimeter in total size (0.9mm3). The MC3571 is only 1.1 x 1.1 x 0.74 mm in size making it 75% smaller than current 2 x 2mm accelerometers on the market today, enabling developers to design high-resolution 3-axis inertial solutions for products that require ultra-small sensor form factors. Our key strength is our speed in delivering both solutions and support to our customers.
mCube Monolithic MEMS (Courtesy of mCube)
YD: We see a clear push from ST in the Chinese market. We also see an increase in market share of your company. What’s your understanding of that?
Ben Lee: Part of our strategy is to continue to increase our market share in China. I think we have a lot of opportunity to do that. We see ourselves as potentially number two in the accelerometer market in China, maybe not this year but certainly by next year. I think eventually we’ll be number one. If you look at the System Plus reverse-engineering report it’s a matter of time until existing players cannot compete with us unless they come up with a brand-new platform that’s better than MCube’s. The size and value advantage are so overwhelming that while some other suppliers are selling at cost and sometimes below, we can still be competitive where we are. We’ll also continue to drive the technology for our miniaturization. Our focus is definitely to continue to widen the gap in our technology leadership. We don’t see any other supplier closing in, in the area of the monolithic single-die accelerometer. As a matter of fact we see existing players moving away from accelerometers – as they’re not making any money from it – to other things, like actuators, and bio sensors. I think in this game it’s the last man standing, but the good news is that the world needs a lot of accelerometers, and somebody has to supply them, so hopefully that’s going to be MCube. Then with that momentum in place we also see a lot of opportunity to do combinationproducts, either by ourselves or with partners using our very small accelerometer as a building block. We’re working with several companies right now to do sensor integration, to do system-on-chip integration. We have more of these projects than we can handle right now. People are coming to us because we are the only company that can supply the single-die MEMS+CMOS accelerometer for integration. I think that MCube could be the ‘MCube inside’ for the inertial sensor market.
Interview realised bu Jean-Christophe Eloy, CEO and President of Yole Développement
Ben Lee is CEO of mCube, Inc. He has over 20 years of senior management experience in the semiconductor industry, with more than 10 years serving as an expat in Asia (Singapore, Hong Kong, and Shanghai). Ben is especially interested in working with innovative IoT solution companies that can benefit from mCube’s world’s smallest MEMS inertial motion sensors. Prior to mCube, he served as Senior Vice President of Worldwide Sales at Cypress Semiconductor Corporation. Prior to Cypress, he served in senior management roles at Trident Microsystems, Apexone Microelectronics, Altera Corporation, National Semiconductor Corporation and Chartered Semiconductor Manufacturing. Ben began his career as a System Integration Engineer at IBM’s Federal Systems Division in New York. He holds a BSEE from California Polytechnic State University, San Luis Obispo, and a MBA from Golden Gate University, San Francisco. Mr. Lee serves on the Industry Advisory Board for Cal Poly Electrical Engineering and is a Board Director of Ten Degrees, Inc.
Go further with our analysis – related reports:
Related Reports and Monitors
TDK-Chirp Ultrasonic Time of Flight Sensor
Reverse Costing - Structural, Process & Cost Report
Consumer MEMS Microphones Comparison 2020
Reverse Costing - Structural, Process & Cost Report