As part of the quarter rollout of new products, Sun's freshly minted president, Jonathan Schwartz, declared that Sun is leading a new era in which the network is a commodity, but Sun's products aren't. CEO Scott McNealy talked about Sun bringing forth a wave of disruptive innovation. "Disruptive" applied to innovation (and borrowed from people like Intel Chairman Andrew Grove and Harvard professor Clayton Christensen) means technologies, like the 8088 microprocessor, or strategies, like Dell's direct sales model, that redefine the competitive landscape.

McNealy wants to turn his company's investments in R&D into disruptions that benefit customers and stockholders, and cause pain for competitors. It remains to be seen whether Sun will succeed in profoundly altering the computing landscape in this century, but the company is clearly testing out some new concepts and taking aim at IBM. McNealy rehashed his "mankind versus IBM" line.

Now that Sun has mostly buried the hatchet with Microsoft, McNealy's verbal barbs are targeting IBM's product family and global services organization.

The major announcement of the day will provide more competition for IBM's mainframes and high end servers. Sun and Fujitsu will merge their Sparc processor-based server product lines by mid-2006. The new servers, code-named Advanced Product Line (APL), will replace Fujitsu's Primepower servers and Sun's Sun Fire line of servers. The R&D partnership leverages Fujitsu's mainframe expertise and Sun's network computing systems design capabilities, and both companies will manufacture and sell the systems. McNealy said that competing against Futjitsu in selling the systems is fine as long as IBM doesn't win the deals they compete for. The "next-generation APL microprocessor...will bring new meaning to the term RISC mainframe," McNealy said during a press conference.

Working with Fujitsu isn't exactly disruptive, but it's smart in that it saves Sun R&D dollars, extends the life of Sparc (which seems be a sacred cow) and could provide the two companies with a strong alternative to IBM's high end systems (especially now that Sun has killed plans for the UltraSparc V and Gemini chips). However, two years is a long time: Who knows what can happen during the development process? An agreement is in place, but detailed specifics about the new Sparc processors, server products and related technologies weren't forthcoming. McNealy did allow the two companies would deliver servers with from 1 to 128 processors and maintain binary compatibility with the existing base of applications. McNealy also boasted that Sun would become the leading provider of scale-out, AMD Opteron-based systems.

Sun is also trying to disrupt conventional pricing models. A new storage solution foray into subscription pricing offers customers a high-end Sun StorEdge 9980 system for $1.95 per gigabyte per month with a 36-month commitment and minimum of 30 terabytes of storage. The base price includes storage management software and support services, but increases by $1.50 for enhanced management or by $2 for mirror-image storage capabilities. The StorEdge 9000 series is geared toward block storage, such as databases.

"It's a risk sharing model, and allows customers to get in without big up-front cost," according to Andy Ingram, vice president of Sun's scalable systems group. "By standardizing things, we can make the hardware acquisition competitive and save on the total cost of ownership." According to Ingram, Sun is already working with partner Affiliated Computer Services (ACS) to provide utility computing services. Either partners or Sun can host the storage utility, or customers can co-host on their own premises.

Sun plans to expand its utility service to include servers, networks, management and applications. "Our goal is to continue to move up the value chain," Ingram said. However, utility-based services that are more complex than delivering block storage or server cycles require far more integration and customization. "We have reference architectures that offer a basic recipe--we can cater the whole meal or they can do it themselves," Ingram said.

Reference architectures, which all the major vendors offer in some form, provide best practices and blueprints for implementing applications such as Oracle 10g, as well as solutions like RFID, VoIP and Grid instrastructure.

I asked McNealy what was disruptive about Sun's utility services model. He characterized Sun's utility as simple and standardized compared to IBM's utility services. "IBM wants to do something custom for you. We give you only one configuration. [IBM Global Services] is a spigot for customer invoices rather than a standard service."

Sun and its partners taking on more complex services--business process outsourcing as well as utility computing--in the data center begins to sound more like what arch rival IBM does with its service group. "It's an issue of how to make it work for all kinds of partners and sales people--we have to learn as we go," Ingram said.

Of course, IBM has a major size advantage. IBM Global Services recorded nearly $100 billion in revenue for IT services and consulting in 2003, and has about 180,000 employees in more than 160 countries. No matter how McNealy characterizes IBM, he can't make it go away by disruptive innovation. That's like Scott McNealy trying to prevent Shaquille O'Neal from scoring.

However, Sun is making progress. The company is the best of its breed at simplifying IT acquisition. For example, the company introduced a collection of more than 100 subscription-based services designed to improve the operation of Sun equipment in a data center. Sun's Preventive Services offers financial incentives of up to 20 percent off long-term service contracts to customers who meet or exceed agreed upon performance goals.

At $100 per employee per year, the Java Enterprise System (which includes the Sun ONE application server, an LDAP-compliant directory server, portal, messsaging and calendaring servers, and Sun's clustering system) is hard to beat as a pre-packaged enterprise infrastructure on Solaris and Linux, with Windows and HP-UX releases upcoming. In addition, Sun is making JES available on a per-citizen basis for federal, state and local governments of less developed nations; pricing ranges from $0.33 to $1.95 per citizen per year. Sun has also extended the current $50 price per employee for the evolving Java Desktop System (JDS) until December of this year.

McNealy wants to lead Sun into places that competitors can't go, with services and subscriptions and shared investments in technologies like Sparc processors. "I am not looking for check but for checkmate with competitors," he said.

As software and hardware infrastructure become more commoditized, in terms of requisite services, Sun could be in a good position to capitalize on the trend, especially among its installed base. Checkmating competitors, however, will take more strategic moves than Sun has so far revealed. A good indicator will be how successful Sun is with JDS versus Microsoft Office and its new suite of identity management products and services versus a host of competitors. Partnering with Microsoft and Fujitsu is a good move, but still a bit of an unknown.

When asked during a press conference to define the metrics of success for disruptive innovation, McNealy replied that the leading indicator would be server unit volume, followed by recurring revenue from subscription services, cash generation (Sun has $7.5 billion socked away) and the vitality of communities, such as developers and members of the Java Community Process. He also noted that for the first time in 22 years Sun was focused on getting its cost structure down.

When McNealy can say that server sales are no longer the leading success indicator, Sun will have done something truly innovative and disruptive.

You can write to me at dan.farber@cnet.com. If you're looking for my commentaries on other IT topics, check the archives.