Analyst Opinion - We had a major reorganization at AMD last week that should not only make them more efficient. It may also make them the leading change agent in the server space as we move to a more cloud-based computing model. However, that begs a change on the client as well and interestingly enough, I was at Intel this week for an update on their Digital Home initiative and think this is the best effort I have seen from them in over a decade.
What is amazing is that if Intel takes what they are doing in the digital home and apply it broadly to the PC architecture, they could create what may be the perfect client for the server that AMD's change may envision. Let's talk about big changes in the PC space.
OnLive: The first clue
What put me in the mindset to be looking for change was the roll out of OnLive, which is an On-Line gaming service promising to provide full scale game system experiences from a cloud based service. The problem with OnLive is that neither the servers nor PCs of today really lend themselves to the kind of intensive video experience that we expect for gaming. This increases the cost associated with bringing out a service like this and slows its adoption significantly.
However, the advantages of paying for just the performance you need on a subscription basis are compelling along with the idea of having an appliance that would work more like a TV than a PC currently does. In other words, you simply turn it on and it works.
Since OnLive, I've been looking for changes in Intel and AMD that might indicate that the firms were positioning against this opportunity.
AMD: Building a graphics server
Servers tend to suck at graphics because they are CPU intensive, generally designed for lots of IO and data traffic, but generally don't even need a graphics chip. The problem is that if you want to virtualize a desktop, you need to provide both GPU and CPU capability. Historically, doing graphics on a server was problematic, because the network couldn't handle the video traffic. But with the move to providing multiple HD streams, the network is being modified by folks like Cisco and their Visual Networking initiative to handle this kind of traffic. So we are only waiting for someone to build a server technology that can embrace the graphics side of the equation.
Well it is certainly interesting that Randy Allen, one of the most respected men in AMD and a guy solidly focused on servers, left and was replaced by Rick Bergman, who had been leading ATI. Hello! The second most powerful guy in AMD is now the graphics guy? This suggests more than just a major shift in the executive team. There may be a coming server part that could embrace the performance of technologies like OpenGL and give us the back end that services like OnLive need to become a mass market reality.
The rest of AMD's reorganization was by the book, eliminating much of the complexity that was slowing their execution, but now that they appear to be moving to a more graphics oriented strategy, bigger changes are likely coming much sooner than otherwise might have been the case. But what about the desktop? Enter Intel Digital Home.
Intel getting it right inside
One of the amazing things about CE companies is that they don't do great customer research. If you've ever wondered why there has never been a solid competitor to the iPod, that's the reason. They generally don't even get why they aren't doing well, because a product idea is created in engineering, tossed over the fence to marketing, and little or no customer consideration is brought in to corrupt what otherwise might be a successful process. Watching some of these companies is like watching a child run with scissors. You know it won't end well.
This is Intel's third run at the digital home and their first two attempts were rather pathetic. Initially, they tried to build their own consumer products and that didn't work and then they came up with Viiv, an initiative that made many of us wonder if the firm had institutionalized failure. But this time, they actually seem to have it right, they have done the research, they have defined a unique platform tightly targeted at the needs of the consumer (both price and performance) and it is arguably the best effort of its type I have seen this decade. What they may have also done, unintentionally, is redesigned the PC.
Intel realized that consumers wanted a visually intense appliance-like experience where the web was integrated, not broken out into a discrete browser. For instance, if you want to know the name of an actor you don't want to disturb your viewing by opening a browser to search for him or her. You just want the name to pop up for a second. Unlike a typical PC that prioritizes computing power, this kind of a system would prioritize video performance and their platform has the potential, with regard to streaming video of doing 3x what even the most high performance PC might be capable of. Think multiple HD 3D streams, compelling menus showing live content, and video texturing. This platform would make services like OnLive not only meet PC Gaming performance requirements, but potentially significantly exceed them.
With price targets starting at set top box levels, think sub $200, this could provide levels of performance when coupled with services like OnLive at nearly unbelievable price levels and potentially create the desktop foundation for this cloud computing world the market appears to be headed for.
Wrapping up and timing
You'll start seeingthe set top box part of Intel's strategy in the fourth quarter and this will likely be the first time you'll see something in this class truly impress the market. Prices will drop sharply in 2010 and this is when you're likely to consider one of the products that use this technology yourselves. The most compelling products will likely be the Blu-ray players that use this new Intel technology, because they will do things like true fast forward, effective up-scaling, and web integration that blow away players currently at the very high end of the market.
Over the next 3 to 5 years, expect the services and PCs to go through a massive change as they embrace this new digital world and the PCs of tomorrow become the appliances we always hoped they would be. It is interesting to note that I'm getting some initial indicators that Apple may actually be the first of the PC vendors to embrace this change while Cisco may be the first server vendor.
In the end it will be a combination of servers tuned to provide a whole new class of high definition visual services coupled with a visual computing network and the revolutionized desktop that will likely come together to create this new world of computing. Something is being born from the fires of this crumbling market and it may have the potential to be truly amazing.
Rob Enderle is one of the last Inquiry Analysts. Inquiry Analysts are paid to stay up to date on current events and identify trends and either explain the trends or make suggestions, tactical and strategic, on how to best take advantage of them. Currently he provides his services to most of the major technology and media companies.
The opinions expressed in this commentary are solely those of the writer.