ARM-based Servers

by mjfern on March 15, 2011

In several previous posts (“The End of x86“ and “ARM Disrupting Intel with its Business Model“) I examined the potential of ARM-based processors to disrupt the x86 architecture. For a disruption of x86 to take place, ARM technology must progress up-market beyond the embedded and mobile markets, into higher performance computing devices, such as servers, where the x86 architecture dominates. To assess ARM’s progression into the server market, I have organized the following collection of recent and relevant articles:

Intel’s dominance in the global microprocessor market is being threatened by the rapid rise of ARM Holdings. With its roots in embedded systems and low-end mobile phones, ARM is fast becoming the processor of choice in smartphones and tablets. In addition, a range of companies are now experimenting with ARM-based processors for larger computing devices, from netbooks to servers (e.g., see NVIDIA with Project Denver).

One of the reasons that ARM is gaining share in the processor market boils down to technology. ARM-based chips are more energy efficient than Intel’s. As noted by DataRespons, an “ARM-based system typically uses as little as 2 watts, whereas a fully optimized Intel Atom solution uses 5 or 6 watts.” Energy efficient processors are well suited for portable devices, where battery life is a primary concern. Energy efficiency is also becoming an important issue for non-portable devices, such as servers, because of rising energy costs and concerns about global warming.

Intel is acutely aware of this technological threat posed by ARM, and is aggressively working on developing more energy-efficient processors. Given Intel’s significant technological expertise, the company should be able to close the energy efficiency gap with ARM, if not in 2011 then certainly by 2012.

But what if the threat posed by ARM is not purely, or even primarily, technological? In other words, what if Intel is able to match or even beat ARM-based processors from a technical standpoint (e.g., energy efficiency and clock speed), but continues to lose market share to ARM in all key segments of computing?

Looking beyond technology, ARM Holdings has several other significant advantages that Intel has yet to publicly confront. Each of these other advantages stem from ARM’s unique business model. Unlike Intel, ARM does not manufacture processors and the integrated chips (e.g., system-on-a-chip). Instead, it licenses its processor technology to OEMs who then incorporate this technology into integrated chip designs. The chip designs are then often produced through a contract foundry (e.g., TSMC).

This unique business model gives ARM three significant advantages over Intel. First, with this model, “ARM offers [OEMs] a considerably cheaper total solution than [Intel] can at present…” (DataRespons). Second, this model enables OEMs to customize integrated chips that conform to different form factors. Customization is a major issue for OEMs trying to shave thickness and weight from the latest portable devices, such as smartphones and tablets. Third, this business model, and ARM in general, is less threatening to OEMs. ARM’s business model gives OEMs much more control over design and manufacturing. Moreover, ARM is not weighed down by the baggage that Intel carries as the only producer of x86 processors used in personal computers (PC), aside from AMD.

Intel’s slow response to ARM’s business model could be a significant blind spot. Or it could mean that Intel is unwilling to directly deal with the threat because any serious response would lead to a significant decline in the company’s revenues and profits. Consider that because of ARM’s unique business model, the company generated revenue and operating income in 2010 that was just 1.0% and 0.7%, respectively, of Intel’s. [1]

Intel is still the dominant producer of processors on a global basis, with “86.4% of the [netbook/laptop] market, 93% of the server and workstation market, and 72.1% of the desktop chip market.” The company has significant financial resources and technology expertise to confront competitive challengers. That said, if Intel wants to retain its command over the processor market into the future, as computing shifts away from PCs to connected devices and cloud computing, the company must quickly and decisively confront the significant business model challenge posed by ARM.

[1] According to Google Finance, ARM’s revenues and operating income in 2010 were $406.6m and $106.96m, respectively. Contrast this with Intel’s revenues and operating income over a similar period, which were $43.6b and $15.463b, respectively.

Facebook’s Future

by mjfern on March 9, 2011

On Sunday night, Warner Bros. announced that it would become the first Hollywood studio to rent digital movies through Facebook. This move suggests Facebook is working on a broader strategy to further penetrate the media distribution space.  Such a move would be a natural extension for Facebook, given it already stores and shares “more than 30 billion pieces of content (web links, news stories, blog posts, photo albums, etc.)…each month.” This static content is in addition to the thousands of apps that are available via the Facebook Platform.

Meanwhile, just over the weekend, there were reports that Facebook is resuming previous talks with Skype to offer online video calls. This would further extend Facebook into the communication space, with a steady progression from wall posts and the news feed, to online chat, to the recently announced “messages” feature, which combines texts, chat and email. Facebook’s communication network is quickly expanding beyond the confines of the company, through the “Like” button, which allows users to post media directly from many third-party websites.

As Facebook is expanding its service in the directions of media distribution and communication, it’s also expanding further into commerce, with the (beta) introduction of Facebook Credits in 2010. Although Facebook Credits are currently used primarily to purchase items in Facebook apps, it’s likely that the Credits system will soon evolve into a full-blown payment platform to conduct transactions for a range of digital products. Facebook Credits can already be purchased via gift cards at a variety of retailers, from Target to Tesco.

To summarize, it appears that Facebook is advancing in three core directions: media distribution, communication, and commerce. As Facebook moves forward on these three fronts, it will create new competition for the likes of Amazon and Netflix. It will also intensify Facebook’s rivalry with Apple and Google, since these two companies are also aggressively pushing forward in similar areas of technology. Because each of these emerging competitors, from Amazon to Facebook, have roots in very different areas of technology, from online commerce to social networking, we should expect a very dynamic competitive environment over the next 3-5 years. This is good news for online consumers, and bad news for legacy firms and industries.

Latest comScore Data Spells Trouble for WP7

March 7, 2011

comScore Reports January 2011 U.S. Mobile Subscriber Market Share The latest data from comScore spells significant trouble for Microsoft in the smartphone market. Despite launching Windows Phone 7 (WP7) with great fanfare in November 2010, its U.S. market share in smartphones has in fact declined 1.7% from October 2010 through January 2011. To regain its [...]

Read the full article →

Rivals Can’t Beat iPad’s Costs

March 7, 2011

A New York Times article appeared on Sunday with the title: “So Far Rivals Can’t Beat iPad’s Price.” The iPad 2, unveiled on Wednesday, offers several sleek improvements over its predecessor. But its most attractive feature is perhaps the same one its predecessor had: the price tag. And what makes that feature even more compelling is [...]

Read the full article →

Will Competition in Search Echo the Browser Wars?

March 5, 2011

Will competition in search echo the browser wars between Internet Explorer, Netscape, Firefox, Chrome, Safari, and Opera? The short answer is no. There are tremendous learning effects, economies of scale, and network effects in search that are not present to the same extent, or at all, in the browser market. Let me explain. Learning effects [...]

Read the full article →

Learning Curves in Tablets

March 3, 2011

This past week included two significant events in the emergence of a tablet market. Motorola launched the Xoom last Thursday, the first credible 10” tablet based on Android 3.0, a newly baked Android operating system (OS) optimized for tablets and larger screens. Meanwhile, just yesterday, Apple announced the second iteration of its tablet, the iPad [...]

Read the full article →

Innovations in UI Design

March 2, 2011

Tobii Technology Unveils Eye-Tracking Device to Move Cursors A Swedish technology company is demonstrating new technology at the CeBIT technology trade show in Hannover, Germany today that uses eyeballs rather than a mouse to move cursors on a computer screen, meaning that users can scroll down a Web page or kill bad guys in a [...]

Read the full article →

The Risk of App Development for Facebook and iOS

February 28, 2011

The latest technology platforms, including Facebook, iOS, and Android, have given rise to a vibrant market of third-party apps. This follows the pattern of application development for other popular technology platforms, such as Microsoft Windows. While the basic pattern of application development may seem familiar, there are at least two major differences in the relationship [...]

Read the full article →

The Future of Barnes & Noble

February 22, 2011

Barnes & Noble: Will It Follow the Borders Plot? While initial analysis would assume Barnes & Noble will benefit from the bankruptcy of Borders, this may not be the case… Barnes & Noble said it is suspending its 25-cent quarterly dividend to preserve its cash to invest in growing its e-reader business, and declined to [...]

Read the full article →