ARM Community: Cortex-M Series – spectacular growth and progress - ARM Community

The IP business is a very competitive world and arguably the one area more competitive than any other is IP for the deeply embedded market. Here numerous processor architectures and IP vendors are each competing for their slice of the pie. However, the fact is that the boat sailed on a new architecture for products such as MCUs (and now mixed-signal devices) a few years ago and that boat is ARM. That isn’t to say we are complacent of course and we have lots of new technology in development to keep our customers’ product roadmaps expanding with new processors from ARM.

My team drives product marketing for the processors that are most frequently found in MCUs and other embedded applications, the Cortex-M series, and we have been doing that since 2008. The first member of the series, the Cortex-M3 processor, was launched in 2005 and it was just getting into production with its first customers as my team took over. Back then we had a vision of bringing an MCU-friendly processor family to the market and then working with our semiconductor partners, tools partners, distributors and middleware vendors to establish a new standard in MCUs. I’d like to think we are succeeding in that; I’ll let you be the judge though.

The first Cortex-M processor-based MCUs started to sell in the market in 2006, modestly at first as our first silicon partners sampled their first chips and won their first designs. 2007 saw the first serious volume production and since then we have gone from strength to strength, faster than anyone could have predicted. The rate of adoption, investment and simply the energy and enthusiasm from our semiconductor partners (and it is a partnership) has been amazing and gives a real buzz to the work.

Let’s look at some of the numbers because they are really the thing that shows where the activity is and how fast our boat is accelerating.

First the number of licenses; a few months ago we passed our one hundredth Cortex-M processor license and nearly 50 of those were for the smallest, the Cortex-M0 processor. That is telling in of itself, the Cortex-M0 processor gives us a solid 32-bit story at a cost point that only 8-bit could previously have hit. In the first half of 2011 we closed 29 Cortex-M class licenses, similar to the number that I think one of my competitors has in total after quite a few years of promotion.

The next one is volume and number of end devices. There are now so many that it is getting difficult to keep track of the total but as far as we can see there are over one thousand Cortex-M MCU parts out there and growing.

The third is simple proof through volume. In 2010 our partners shipped around 140 million Cortex-M MCUs and smart cards, and that doesn’t even count the ARM7 processor-based devices which are in similar quantities and still themselves growing.

I’ve not yet used this graph publically but we are quite proud of it and it is a good chance to show it off:



And as I mentioned earlier, this is in addition to the large volumes of ARM7 processor-based MCUs that are already sold every year.

Something that is even more encouraging is the fact that less than ten of our Cortex-M licensees are contributing to this volume, more than 70 other companies have licenses and are still busy developing their products. One great example of that is Freescale with its Kinetis family of MCUs. They have shipped more than 30,000 samples and 15,000 development kits with many parts starting their production ramp over the coming months; given that the interest in these products has been huge we are looking forward to when that happens!

The other markets for Cortex-M
I’ve talked a lot about MCUs so far and the rapid ramp in production of these (to date over 240M of them!) but that isn’t the whole story by any means. Adoption into many other markets where a small, energy efficient processor is needed is also very healthy. If I look at the whole of Cortex-M shipments across all markets we are well over 200M per quarter now and growing rapidly. In fact looking at the data one of our partners will ship the one billionth unit of Cortex-M this quarter!

Crossing over into new markets
In addition to the above, the Cortex-M is also proving to be the “cross over” technology that software developers have long hoped for, with many of our licenses for the System-on-Chip and the mixed-signal markets. For the first time a programmer can move their software between a project that uses a catalogue MCU, a project that uses a highly integrated ASSP or ASIC and another with a true mixed-signal chip, all integrating one or more Cortex-M processors. This cross-over has never been possible before. The traditional MCU architectures have rarely (if ever) been able to break into other spaces, with the possible exception of the 8051. Particularly now with the Cortex-M0 processor there is no reason why anyone would need to choose an 8051; a Cortex-M0 processor-based system will be smaller, lower power and more efficient!

I am particularly proud about our wins in the very solid mixed-signal market. Some of those are set to start shipping in 2012, giving another great burst of success and we are really looking forward to being able to talk about them.

Conclusions?
What conclusions should you draw from all of this? Well the main and the most important one is that ultimately market adoption and use in products that are selling in huge volume today is the real indicator of product quality, stability, and ecosystem support; if you’re not designing with ARM Cortex-M processors, you need to ask yourself if it is time to think about changing? Me and my team would love to talk to you if you are, it doesn’t matter how small or large the opportunity, to us every design win counts.

Richard York, Director of Embedded Processor Products in the ARM processor division with responsibility for the team marketing ARM’s embedded and microcontroller CPU products including the Cortex-M and Cortex-R series. He is also responsible for the overall embedded roadmap for these products and the ecosystem around them and works closely with ARM’s licensees and their customers as they develop their products and markets. He has worked at ARM since 1994, during which time he has been closely involved with the design of the early ARM processors before moving into a marketing role in 2000.