In addition to the keynote at the Internet Global Congress, which I wrote about in an earlier post, I participated in a Power.org conference in Barcelona. Power.org is an initiative launched by IBM last year to bring open, collaborative innovation to the world of systems hardware and microprocessor design and is centered around the Power architecture. We launched Power.org in Beijing last December, at which time we announced the first round of partners. In Barcelona last week, we announced additional Power.org partners, as well as new products and other developments from the Power.org community.
Why is it important to have the kind of open, collaborative innovation in the world of hardware that we are fostering with Power.org and that we already have in software? When all is said and done, the key to successful open communities is that they allow all their participants to innovate, especially customers and users. This is the breakthrough with open source software that Power.org seeks to emulate at the hardware level.
As technologies become increasingly powerful, inexpensive, and ubiquitous, the technology itself is much less important than the innovative products and services it helps us create. This has been true through history; think of electricity, gasoline engines and telecommunications as technologies that have pretty much "disappeared into the woodwork." Such a disappearing act is the ultimate compliment you can give a technology. It means that the products and services that use it are so well designed, that the underlying technologies do not show through. A well designed car is beautiful to look at, a pleasure to drive, makes us feel comfortable and, of course . . . the technologies also work exceptionally well.
With few exceptions, the products and services we have created in the IT industry are far better known for frustrating their users than for exhibiting elegant designs. I believe this is starting to change, both because of the huge recent advances in performance, price, quality and standardization of IT components, as well as the desire to reach many new users with innovative products and services. But such elegant designs cannot be achieved by the IT industry alone. More than ever, we need partnerships between the experts in the underlying IT components and the experts in the products we want to support with these components. Nowhere is this more important than in the emerging world of ubiquitous or pervasive computing, where we are almost literally trying to integrate computing into the physical world all around us, everything from entertainment and appliances in the home to medical equipment and cars.
We think that the Power family of microprocessors is an excellent base around which to build many of these new products and services. But to do this, we need a very close collaboration between our people in IBM, who are experts in microprocessors and related technologies, and our partners from different industries, who are the experts in their products. They may need extensions to Power to better adapt it to whatever they are designing, and, rather than having to serialize their requirements by giving only IBM people access to the Power specifications, we took a lesson from software's open source communities and are making available the Power specifications and tools so that our partners can make the needed changes themselves.
The Cell processor is a very good way to illustrate this point. The Cell processor was originally designed by IBM, Sony and Toshiba as the micro for the Play Station 3. Power is the underlying architecture of the Cell processor, but Cell includes a number of special accelerators that are particularly effective for visual and real-time applications such as those used in games, and for those kinds of applications Cell is incredibly fast. We already know that quite a number of other applications in science, engineering, health care, and education can take advantage of Cell-like capabilities, so we plan to make Cell more generally available in the future. In fact, to further enable innovation in Cell applications, we just announced that we are releasing the Cell specifications so others can design new products around it.
IT is being integrated increasingly into all aspects of business, society and our personal lives. But, to be able truly to take advantage of IT and enjoy information technologies, we will need to "humanize" the technologies by designing products and services that really adapt themselves to us, rather than the other way around. This is a major objective of the collaborative innovation exemplified by Power.org.
"Humanize" is the right word and the Cell processor might be a good opportunity start doing it. Humanizing also means making something easier for the user and that's exactly what I try to achieve with wyoGuide ( http://wyoguide.sf.net ) and wyoDesktop ( http://wyodesktop.sf.net ).
While wyoGuide shows "How to build well-designed applications" it also shows "How to do it cross-platform", so these humanized applications not only run on Windows/Intel but also on the PowerPC. IMO the Power.org is well advised to look into wyoGuide even if they don't care much about humanize. But do they recognize what wyoGuide means for them, how they could profit from it?
wyoDesktop is on another level, it's a dream, a vision of a easy usable free desktop system, albeit a vision with a sound realization plan. Yet again the Cell processor would be an ideal partner for such a project, but does the Power.org have enough courage to size this opportunity? Does the Power.org have enough vision to see the value of wyoDesktop even if it's in a rather early state?
To sell a car it needs more than just a superb engine, it also needs a nice form, comfortable seats, a easy wheel and understandable instruments. The Cell is just the engine while wyoDesktop is a plan to build this engine into a full exceptional car and wyoGuide takes care that this car can be driven by drivers of any country.
The question is now, does the Power.org wants to sell engines or does it wants to sell cars? Keep in mind I don't mean cars like the Ford T-model (as is the current software state of the art), I mean an up-to-date Ferrari.
Posted by: Otto Wyss | June 21, 2005 at 09:16 AM
Otto - you make a good point that humanizing technologies requires collaborative innovation in the underlying user interface technologies. The "superb engine" in fact becomes an additional motivator for the "nice form, comfortable seats", etc - the nicer/better/more suited one piece of the stack is, the more glaring the issues with the others.
As far as power.org the organization goes - think of it as like eclipse.org (for example) in that the organization itself has nothing really to sell... Its a community of different companies and individuals figuring out which areas they need to collaborate in. As yet, the community is pretty young so we don't yet have the part of the power.org site that facilitates this, but it should be up shortly. If you're willing to forbear until it is, I'd suggest resurrecting this thread in that forum when its live... There have been a lot of thoughts along this line, and I for one at least haven't been aware of wyoGuide and wyoDesktop. There is sure to be a lot of interest.
Posted by: Nigel Beck | June 22, 2005 at 06:14 PM
Hi Nigel, it's nice to hear that there is interest. It would be perfect if this fine computer Ferrari one day can be bought in a shop by anybody and used without requiring to be a formel-one pilot but just a normal driver. I think most needed parts are available today and only need to be put together in a orderly fashion. But this putting together is a rather huge work (split into many different tasks) and will need quite some time, time we all probably have least. Still I think with my 2 projects I've a good base and I have a reasonable realistic plan in mind how this could be made real.
I know the Power.org is young and first has to set up the facilities but there are some time lines ahead which shouldn't be missed. The first date is the O'Reilly European Open Source Convention (October 17-20, 2005 - Amsterdam, http://conferences.oreillynet.com/eurooscon/ ) where I try to talk about wyoGuide. See, regardless how good wyoGuide is, it only will be successful if others know it and of course use it. So the first step is to make wyoGuide known to a broad audience and this is IMO best possible by talking at conferences.
My plan to make this all real is based on a full chain of steps and after each step the plan can be reevaluated and readjusted or even dropped completely. But the first step is the O'Reilly conference and this step has to be clear about now.
Posted by: Otto Wyss | June 23, 2005 at 06:39 AM