ARM Holdings(ARMH), the British designer of microprocessors whose technology is found in almost all mobile computing devices, announced at its recent analyst day more aggressive plans to take share in the server market. Investors should welcome the announcement, as the company seeks to leverage its technology into a market where it has virtually no presence. But has something changed to substantiate the bold claims, or is it just all talk?
Servers: the next frontier
ARM Holdings is proud to show off its conquest of the mobile computing market, and rightly so. The company has at least 90% or more market share in smartphones, wearables, storage, and automotive infotainment, as well as at least an 85% share in tablets. Mobile devices (tablets, smartphones, and laptops) have been growing at a blistering pace of 40% per year from 2009 through 2014.
But according to its analyst day presentation, that growth is expected to slow to a paltry 7% from 2014 through 2020, as the markets for these devices become increasingly saturated. This is exactly why the plan to move into the untapped server market is a great idea to find more growth.
History has shown that ARM Holdings is very capable of diversifying and moving into related markets. In 2004, it shipped 69% of its ARM-based chips to mobile devices and only 5% to embedded computers. By 2014, mobile devices were only 45% of chip shipments, while embedded increased to 34%. Adding servers will offer another revenue stream and market opportunity.
The optimistic case for ARM servers
The company currently holds less than 1% of the server market, but at its analyst day, executives announced they are targeting a 25% share by 2020. This was a notable increase from their previous 20% target, so what has changed to make them more optimistic?
First, ARM Holdings has been increasing R&D and engineering efforts to support non-mobile computing. It has acquired nine companies or teams and increased engineering headcount over the past few years, and it expects the number of non-mobile engineers to nearly equal mobile engineers by 2016.
Second, the company is looking to take advantage of what makes ARM-based chips great and bring these qualities to servers. Intel is known for its cutting-edge chips that are great for high-performance and general computing needs. But ARM-based chips are valued for their efficiency, lower power consumption, low cost, and the ability to embed and integrate them into devices for specific purposes.
Finally, ARM Holdings announced that it currently has six chip vendors in production, with four more expected by 2016. This could be the turning point management is seeing and what caused them to increase their target market share.
The pessimistic case
The industry has been talking about ARM-based servers for a long time now. Years ago, early entrant Calxeda placed its bets on ARM-based servers and sputtered to its demise last year, while both Samsung and NVIDIA have abandoned their ARM-server plans.
One of the critical components that hasn't developed along with ARM server chips is the software. Currently, only Ubuntu Server Linux supports ARM's latest v8 server processors, so more operating systems will need to join to make it attractive. However, ARM Holdings executives did point out that Linaro, an open-source collaborative engineering organization, has been a key contributor to building software to support the ARM architecture. Its members include heavy hitters such as AMD, Texas Instruments, Facebook, Hewlett-Packard, and Red Hat.
Encouraging, but don't hold your breath ...
All of this is encouraging for ARM Holdings, but when you look at the facts, it just doesn't have a lot to show for its progress yet -- management comments were limited to "players are 'evaluating' their offerings" or that the company is in "active engagements" with others.
However, ARM Holdings will continue to be a major beneficiary of mobile computing and increasing Internet of Things applications. The company continues to put up earnings growth of 20% or greater, and because of its licensing model, it has nice, fat profit margins of 50% or more. This makes ARM Holdings a cash machine, with no debt on the balance sheet to boot. Investors who don't mind waiting for additional markets such as servers and the Internet of Things to materialize could do a lot worse than investing in this company.