Jack Gold who has consulted many friends in the industry and provide great analysis now at J Gold Associates has done a more thorough analysis of the Intel Nokia deal that I highlighted yesterday. Having talked to friends in both companies, I think Jack understands the considerations very well. Take a read.
Today, Intel and Nokia announced a long term strategic relationship. There are 3 key parts to this agreement: Intel will license Nokia's technology for 3G HSPA cellular technology for use with its chips; Intel and Nokia will collaborate on making their respective open source implementations of Linux for small devices (Moblin and Maemo) more compatible; and they will collaborate on future Intel architectures (IA) for mobile devices. This is a compelling partnership for several reasons.
First, Intel has had difficulty producing competitive cellular radio chips, going back a number of years to when it had the XScale product line before divesting it to Marvel. Nokia, on the other hand, has had compelling Intellectual Property (IP) in radio technology that clearly made it the world's largest manufacturer of cellular phones. Intel rightly understands that it needs to be able to offer a competitive cellular modem to fill in its communications product offerings (e.g., WiFi, WiMax). This is critical in the netbook and Mobile Internet Device (MID) space where it has targeted its Atom processors, and where it hopes to eventually make a play for smart phones as well with future, lower powered models of Atom. Having 3G HSPA cellular chip competence is therefore critical. Intel would not specify which chip family it would announce products for or when, other than to say it is for the IA set of products. However, we expect the first products using this new capability will be Atom-based systems, and we expect first products to be released in early to mid 2010.
It is interesting to note that Intel and Nokia did not discuss any relationship for 4G cellular technologies (Long Term Evolution, or LTE) which will achieve significant market share in the next 3-5 years. Although we will see some LTE systems in place in 2009/10, we do not expect critical mass to be achieved until 2012 at the earliest. So Intel does have some time to assess what it needs to do. However, Intel does have a stake in a competing 4G technology, WiMax, which it is pushing aggressively. We hope Intel does not assume that it will not need an LTE solution going forward, as we are sure that it will, and it will need one no later than 2010/11 to stay competitive. This will complement, not replace its WiMax offerings. Nokia does have significant IP in 4G as well, so an extension of this relationship in the next couple of years to cover 4G, assuming all works out well, is likely. This means that Intel will not have to depend on third party solutions for cellular chips as it does currently. It also eliminates Intel's dependence on major suppliers like Ericsson for modules and Broadcom and Qualcomm for chips. Qualcomm, in particular, is attempting to move upstream into mobile processing (Snapdragon) which puts it in direct competition with Intel for its Atom processors. This deal therefore eliminates the need for Intel to buy from a direct competitor.
In working with Intel, Nokia gets to have input on the long term evolution of the IA architecture as it relates to wireless communications. This provides Nokia with some important benefits. Nokia is currently dependent on the ARM chip architecture for most of its smart phones and its Internet tablets. Yet it rightly understands that the ARM architecture falls short as it tries to move upstream and impact the netbook and MID market. By working with Intel, Nokia gets to influence the design of Atom chips specifically targeted where Nokia needs to go; expanding from its smart phones base and into more wireless entertainment devices. Using an Atom core built around the IA architecture has great benefits for these higher level devices (e.g., existing application code, programmers and compilers), and Nokia's collaboration with Intel will influence these chips to include more wireless friendly capabilities. We do not envision Nokia abandoning its core dependence on the ARM architecture in the short term, but longer term (2-3 years) we expect Nokia to offer devices based on Atom, especially with System on Chip (SOC) designs. This could provide Intel with a very large marketplace as such devices will be sold in the millions or tens of millions of units per year.
From the Operating System (OS) level, this partnership provides an advantage to both Intel and Nokia. Both have competing Open Source solutions that are built on a Linux kernel, but provide a non compatible user interface. Working together to build more compatibility into Moblin (created by Intel but now Open Sourced, although Intel is still the major contributor) and Maemo powering Nokia's Internet Tablets will have some important effects. First, by making the user interface (UI) compatible, it will allow application providers to build a single application that runs on both Linux distributions. Second, Nokia's expertise in wireless requirements for an OS and UI, together with Intel's expertise in optimizing software for specific chips (with its compiler and newly acquired WindRiver expertise) will make a more compelling and better/faster running system available to the end user. Finally, this collaboration could limit the impact Google's Android OS will have on the netbook market, as Google's stated goal is to make Android netbook and MID friendly. We would expect to see other companies (e.g., Novell, Canonical) support the Intel and Nokia efforts in this regard, as both are already committed to supporting Moblin, and likely see Android as a long term threat.
Bottom Line: This strategic relationship is a win-win for both Intel and Nokia. It is also a win for the marketplace, as it should allow more capable wireless devices to make their way to market in the next couple of years. And a more converged Linux landscape in the MID and Netbook market will allow the market to expand more quickly as fewer incompatibilities in applications and peripherals will exist. We expect to see further relationships develop between Intel and other major stakeholders in significant niche areas of the market where Intel does not have the IP it needs. The days of "do it all ourselves" are over for Intel. And Nokia also understands it needs more partnerships to stay ahead of its competitors
Jack Gold is the founder and principal analyst at J.Gold Associates, an information technology analyst firm based in Northborough, Mass., covering the many aspects of business and consumer computing and emerging technologies.One personal note. I think the reality of the death of CDMA has impacted Intel's WiMAX strategy. Its not that they wont go forward, but supporting LTE is a must for Intel.