Trends, risks and opportunities
The semiconductor industry goes through cycles which include periods of increased consolidation, especially when revenue growth declines. In 2009, there were several examples of ARM’s customer base consolidating either through acquisition of entire companies or one company acquiring the division of another. This trend appears to be particularly marked in the wireless baseband business, where ARM has a large market share and in microcontrollers where ARM is looking to grow its share.
About half of ARM’s revenue comes from direct sales to semiconductor companies. If there are fewer semiconductor companies, then ARM may have fewer customers to sell to. Two-thirds of ARM’s unit shipments come from wireless devices, and some of ARM’s future growth comes from the microcontroller market, so consolidation in these parts of the industry could represent a loss to ARM’s future licensing business. In addition, due to reductions in venture capital funding, there may be fewer technology start-ups to license ARM technology.
Saving cost is often a major driver when two semiconductor companies come together. ARM works with the merged company to demonstrate how increased cost savings may be realised through the adoption of ARM as a common technology across more divisions in the merged company. None of the consolidation we have seen to date has had a meaningful impact on ARM’s licensing opportunity as much of the consolidation consisted of companies selling divisions, where both companies remained ARM licensees after completion of the transaction.
In addition, as about two-thirds of ARM’s licences are signed for non-mobile applications, much of ARM’s licensing opportunity is beyond wireless. With the introduction of Cortex-M class processors ARM has increased its market opportunity in microcontrollers and we have more companies to sell to than ever before, and in 2009 we had 20 companies license their first ever ARM processor.
Smartphones are getting smarter and laptops are getting smaller and more portable, and we are seeing new mobile computing products introduced. This is creating an opportunity for smartphone technology to cross-over into laptops and laptop technology to cross-over into smartphones. Consumers want portable products that keep them connected to their social and business networks, have an all-day battery life and are simple to use.
The main processor in a laptop is typically based on the x86 architecture. It has been announced that smaller and lower-power x86-based chips are being developed that will be suitable for the main processor in a smartphone. There is therefore competition from large, well-funded companies that have demonstrated advanced technology and aggressive marketing strategies. They are capable of reducing ARM’s market share in smartphone application processors and hindering any market share gains that might be made by ARM licensees in mobile computing.
ARM-based chips for mobile computing are significantly lower-cost and lower-power than current products available based on x86. ARM’s licensees have already announced chips that are suitable for mobile computers, and they regard this market as a growth opportunity. The ARM community is working with software companies and OEMs to ensure that a complete ecosystem of PC-class software is available, and that our shared experience in developing portable consumer products enables a new market of low-cost mobile computers.
In addition to the main microprocessor, mobile computers contain many chips that can incorporate ARM-based technology. We expect that most mobile computers could contain 3–5 chips that would increase additional ARM royalty revenues.
Although growing strongly, in 2009 ARM had only about 6% market share of the nine billion unit microcontroller market. This market is characterised by a large number of proprietary processor architectures, being developed by many semiconductor companies, who sell very low-cost chips into a highly fragmented end-market.
It could be difficult for ARM to be successful in the microcontroller market. ARM will need to displace many established in-house processor designs. ARM may invest a lot of effort and cost to achieve modest penetration. As the microcontroller chips are low-cost, the royalty revenue per device will also be lower.
ARM believes that it can capture a significant proportion of the microcontroller (MCU) market as demonstrated by the strong unit growth rate over the last few years. ARM processors are already 30% of the 32-bit MCU market and our strategy is to develop technology that enables companies currently using 8- and 16-bit MCUs to migrate to ARM-based chips. As most semiconductor customers already have access to ARM technology, many have already produced MCUs based on the ARM7 or ARM9 processor families.
The Cortex-M processor family was specifically developed for the MCU market and, by the end of 2009, had been licensed over 50 times. In addition, the ARM community is working to ensure that as well as a broad range of ARM-based chips, all the software and software tools are available for end customers using established microcontroller distributors.
As semiconductor manufacturing has become increasingly expensive, most chip companies are considering whether to outsource their manufacturing facilities to specialist foundries. This creates a new opportunity for chip companies to license in their physical IP R&D. As all chips require physical IP technology, this could be a larger market than processors.
Currently, most major chip companies develop their physical IP using their in-house teams. Even for companies that have outsourced manufacturing, the rate at which they want to license in physical IP is unclear. The foundries may choose to develop the physical IP. This could add more value to their customers and help create ‘lock-in’ by making it harder for the customer to change foundry. There are also other independent physical IP suppliers who compete in this market.
ARM is a proven supplier of processor technology to most of the major chip companies in the world. We have developed the most advanced physical IP technology and we were market leaders in the physical IP market in 2009. Currently this is a small market that we believe will grow over the medium-term as more major semiconductor companies outsource their manufacturing. ARM can demonstrate significant cost savings for our customers if they license physical IP from ARM; and with the combination of advanced physical IP and processor technology we have an unmatched offering.