For some years, the cellphone has been the dominant application for shipments of the ARM architecture, with some industry watchers commenting that ARM was overly dependent on just one application. While historically that analysis may have been valid, recent research from Semicast suggests a major change in the future application make-up of ARM shipments, with a bright future beyond the cellphone.

According to Colin Barnden, principal analyst with Semicast, shipments of ARM-powered products in applications other than cell phones totaled a not inconsiderable 950 million in 2006, but he forecasts that this will grow to 3.3 billion in 2012, a CAGR approaching 25%. Compared to unit growth in cell phones (at a CAGR of only about 8%) it is evident that while ARMs past was about the phone, the future is about, well, pretty much everything else.

Semicast has identified some key factors which will help develop the future of the ARM architecture beyond the cell phone. These include:

The Cortex-M1 Core: First announced in April 2007 and developed by ARM in partnership with Actel specifically for integration into FPGAs. The introduction of Cortex-M1 follows Actels introduction of a soft ARM7 core for its ProASIC3 product line in March 2006. Cortex-M1 will help establish ARMs presence in FPGAs, where it will compete with Alteras Nios and Xilinxs Microblaze cores for design wins.

The Cortex-M3 Core: First announced in March 2006, with ARM working with lead partner Luminary Micro to introduce a range of low cost devices (down to $1), to compete with 8/16-bit MCUs. Cortex-M3 will play a critical role in establishing ARMs presence in the low cost MCU sector.

Support from a Connected Community: In recent years a number of key suppliers have introduced and aggressively marketed a range of ARM-based MCUs, such as Analog Devices, Atmel, NXP, Samsung, STMicroelectronics and TI. It is highly likely that other suppliers will proceed to license the ARM architecture and introduce their own ranges of ARM-based MCUs. This could potentially include suppliers such as Cypress Semiconductor, Infineon, Maxim, Microchip and Zilog.

Development of ASSPs: ASSPs are forecast to continue to dominate shipments of the ARM architecture, accounting for over 90% of units. While cellphone chipsets represent the largest share of ARMs heritage, it is highly likely that a sizable part of the future lies with customized solutions developed by design houses (especially in China and Taiwan) which license the ARM architecture to introduce very low-cost chipsets targeted at specific high growth applications. A good example of the rewards that can be achieved from this design route is Actions Semiconductor, based in Zhuhai, China, which offers a range of low-cost ASSPs for MP3 players. While not an ARM licensee, Actions Semiconductor is one of the leading controller suppliers to Chinese MP3 player manufacturers. While Apple accounts for much of the hype in the portable audio market, shipments of low-cost, low-end MP3 players from Chinese manufactures (which receive little hype or fanfare for their products) far exceed shipments of the iPod. Actions Semiconductor reported revenues of $170 million in 2006, with net income of $74.5 million. At almost 45%, that is a level of profitability generally unheard of in the semiconductor industry and, as a percentage of revenues, its net income exceeds that of Intel.

Life Beyond the Cellphone: Continued pervasion of fast internet access in the home is forecast to drive strong demand for wireless internet hardware and home networking, with strong growth also in storage and printing. The growing acceptance of Skype(tm) and similar services, for voice-over-internet calling is also likely to raise demand for ARM in a new type of phone, the VoIP phone; Broadcom is one supplier offering a range of products for this application. Automotive and industrial applications should not be overlooked, with a combined ARM market approaching 500 million units in 2012.