How serious? Does $2.2 billion and 15 percent EBIT profits sound serious to you? It does to me, and also to Deutsche Bank analyst Chris Whitmore.

Having slogged through Oracle’s 72-page document dump with a better eye for detail than mine, Whitmore noticed a line in a January 2010 email from Dave Donatelli, now head of HP’s Enterprise Group (specifically Exhibit 17, for those who want to scroll through and find it) saying that HP’s Business Criticial Server business combined with its Technology Services business, which includes the support and services associated with the Integrity line of servers that uses the Intel-made Itanium chip, was at that time larger on a revenue basis than HP’s personal computer business.

The same document, he says, showed that at the time, HP’s “owned operating profit” for the combined hardware, software and services tied to the business of selling and supporting Itanium servers was about $2.2 billion. All in, HP derives — or at least at that time derived — about 15 percent of its profits on an EBIT basis from Itanium and related businesses.

No wonder, then, that HP considered Oracle’s March 2011 decision to stop creating software that runs on the Itanium chip so earth-shattering that it hauled the software giant into court last June. That case is expected to head to trial any day now.

The disclosure is the clearest sign yet of how much HP stands to lose if its Business Critical Server business can’t recover. It has always been known to be a highly profitable business; exactly how profitable was a closely guarded HP secret. But sales of Business Critical hardware have been on the decline. In 2009, sales of BCS hardware were $2.6 billion. In 2011, they had fallen by 19 percent to $2.1 billion. And in the quarter ended Jan. 31, sales were $405 million, down 27 percent from the same period in 2011.

The uncertainty about Itanium’s future is one of the many reasons that Whitmore has been particularly bearish on HP’s turnaround prospects: “Given the growing uncertainty around the long-term viability of Itanium, we expect customer defections to continue, if not accelerate in future periods,” he wrote in a research note to clients, issued yesterday.

However, much as HP lawyers would like to argue that Oracle’s motivation is to help bolster long-flagging sales of its new Sun Microsystems hardware unit, Whitmore argues that the main benefactor is IBM: “While Oracle is responsible for shining a bright light on Itanium’s precarious future, it is probably doing IBM the biggest favor. … We expect IBM to be the greatest beneficiary of Itanium defections and view Power [IBM's server chip] as the market consolidator and eventual standard in the UNIX/RISC server market over the medium to longer term.”

And even if HP prevails in its suit, Whitmore isn’t seeing much benefit: “Regardless of the outcome of this particular suit, we expect HP-UX customers to continue fleeing what is increasingly looking like a dead platform — creating a major headwind for HP’s medium-term earnings.” Ouch.

Obviously, that’s not the part that surprises me. But here is the bit that did: When you add up all those expenses over a server’s lifetime, labor costs amount to about 70 percent of the total, according to IBM. If you had asked me, I would have guessed the cost of power would outweigh the cost of ongoing labor. Silly me.

I talked with IBM’s Steve Mills about this earlier this week. He’s Big Blue’s senior vice president and group executive for Software and Systems. It’s not uncommon, he says, for a company to take weeks or even a month between a server’s arrival and its deployment.

IBM today announced a hardware system it calls PureSystems that can cut that deployment time to hours and reduce the lifetime labor cost associated with the server by about one-third.

Basically what IBM is doing here is bringing to bear its expertise in services. Having done so well running IT services for a few thousand different companies, it has learned a thing or two about efficiency.

And it makes perfect sense when you consider that much of IBM’s $107 billion in revenue is derived from its services business. Now it’s taking some of that learning and applying it to its hardware and software business, which accounts for about 40 percent of sales.

The key feature, Mills told me, is something called the Flex Systems Manager, which is some IBM-made software that automates a lot of the set-up and maintenance work that traditionally has to be done more or less manually by one or a team of IT managers. “The purpose of the code is to do discovery. … Can I locate every piece of hardware in the frame? What are the rules for configuring it? Can I locate all the software I need and what are the rules for configuring that?” Mills told me.

All that data has been gathered into a single screen that makes the relevant information available at a glance. Mills says the system can be up and running within four hours of arriving at a company’s loading dock. That’s a bold claim.

It’s all based around patterns that IBM has seen over and over again for different types of deployments and configuration options. See them often enough and you can develop software scripts that take a great deal of the manual labor out of the process.

Sometimes companies have their own unique or wonky business processes that even someone as experienced as IBM hasn’t seen before. If that’s the case, a company can craft its own pattern and translate that into software that can automate a process that’s unique to its business or internal rules.

IBM has also teamed up with 125 independent software vendors or ISVs to develop their own patterns that clients can quickly download in order to get up and running. (IBM put out a video on that, which I’ve taken the liberty of embedding below.)

It’s also pretty diverse from a computing standpoint. IBM being IBM, the system has different hardware options, including processors from Intel or its own Power line of chips. There are also three OS options: Windows, Linux and AIX, IBM’s proprietary flavor of Unix. There’s also a wide choice of virtual machine managers: VMWare, KVM, Microsoft’s HyperV and IBM’s own PowerVM.

In the end, the point is to allow a company’s employees to spend more time working on their key lines of business and less time making the computers run properly, which is at its most basic level the IT shop’s highest mission.

One thing Judge James Kleinberg did was dismiss a fraud claim by Oracle that said HP had been all sneaky when it concluded a settlement with Oracle that included an agreement to continue building software for systems using the Itanium chip. The settlement was struck only a few weeks before HP hired Léo Apotheker as its CEO and Ray Lane as its chairman.

But that’s not the important part of what Judge Kleinberg did. The most important aspect of yesterday’s action in Hewlett-Packard v. Oracle was the release of the unredacted version of Oracle’s cross-complaint. And it’s a juicy read.

I’ve posted the redacted version before. Now you can read all the bits that were blacked out.

What you’ll find is a lot of information that goes to the core of Oracle’s argument that HP has a lot to lose if the Itanium chip goes end of life, which is exactly what Oracle has said Intel plans to do. As the only major server vendor who sells boxes running Itanium chips, HP makes a lot of money — billions of dollars, according to a newly unredacted statement in the filing — on service-and-support contracts with its Itanium customers. As one HP executive is quoted on page four of the filing, without Itanium, HP would be “strategically screwed.”

Intel, on the other hand, was more or less ready to let the chip die. Having spent billions, dating back to 1989, to develop the Itanium chip, which outside of HP never saw any market success, Intel had to be convinced to keep building them. To do that, HP, the filing reads, paid Intel $440 million to keep Itanium chips in production for a few more generations, through 2014. The deal didn’t even cover the cost of the chips, as HP had to pay for them, as well, the filing reads. Oracle calls the arrangement a “pure pay-off to induce Intel to keep churning out processors that it really wanted to kill.”

And while there’s nothing specifically wrong with the arrangement by itself, Oracle’s point is that HP was misleading the marketplace about the true status of the keystone product in its Business Critical Service business. That unit, in no small part because of the uncertainty wrought by this lawsuit, saw its sales fall by 23 percent in HP’s most recent quarter.

Having won the release of the unredacted complaint, Oracle claimed something of a victory in a statement:

“Oracle is delighted that the Superior Court of the State of California, Santa Clara County, has rejected HP’s attempt to hide the truth about Itanium’s certain end of life from its customers, partners and own employees. We look forward to seeing all of the facts made public that demonstrate how HP has known for years that Itanium is end of life.”

It all sounds very reasonable, until you take into account the fact that Oracle acquired Sun Microsystems in 2010 and is now a big server vendor that competes with HP, and would by no real stretch of argument benefit from an exodus of HP’s Itanium customers toward other vendors. HP called the decision by Oracle to cease support for Itanium part of a “calculated business strategy” to mess up HP’s Itanium business and capture those customers. Yet the evidence so far suggests that the one benefiting from this fight is actually IBM.

HP claimed victory of its own, in a statement:

“HP is pleased that the Superior Court of the State of California, Santa Clara County, has rejected Oracle’s attempt to use a fraud claim to undo its contract with HP. We look forward to seeing the facts made public that demonstrate how Oracle’s March 2011 announcement to no longer develop software for Itanium servers was part of a calculated business strategy to drive hardware sales from Itanium to inferior Sun servers. This further demonstrates the fact that Oracle breached its contractual commitment to HP and ignored its repeated promises of support to our shared customers.”

HP has portrayed itself as the defender of the interests of Itanium customers, under attack by Oracle. As HP puts it in its statement, Oracle tried to induce customers running Oracle software on HP Itanium systems into replacing that hardware by limiting support and withholding software patches and bug fixes. “Customers were left without options to address bugs and other defects in their Oracle software,” HP says.

For HP, this is all a simple argument over whether or not Oracle can be held to the contract they agreed to in 2010.

The agreement stems from the circumstances of former HP CEO Mark Hurd’s resignation, and his subsequent hiring by Oracle as its president. HP sued Hurd and Oracle, and soon they settled. HP says that a clause in that settlement included a provision that Oracle would continue to port its database software to HP servers running the Itanium chip. Oracle has argued that this clause is not part of the final agreement. The settlement document itself remains confidential, but its details will likely emerge in the trial. Expect lots of arguing over different versions of the agreement.

I have embedded two documents below, for your reading pleasure. The first is Oracle’s unredacted cross-complaint, with all the blacked-out bits from the previous version now fully revealed for the world to see. Below that is a Case Management Conference Statement filed by HP lawyers, also unredacted, where it seeks to expose Oracle as making cold-blooded moves that would appear to be attempts to spur Oracle’s own software customers to abandon HP hardware. It’s not quite as juicy as Oracle’s document, but it has its moments, too. Enjoy them both:

HP v Oracle – Amended Cross Complaint

0077a 2011121 Hp Cmc Stmnt Unredacted

The expected filing is Oracle’s amended cross-complaint, wherein the company will lay out the basis of its legal argument explaining why its March 23 decision to stop building software that runs on servers using Intel’s Itanium chip was not only justified but doesn’t violate an agreement struck last year between Oracle and HP.

Oracle has made several colorful claims in court. Last month, for example, it compared an arrangement between HP and Intel to continue to produce and evolve the Itanium chip to a remake of the film “Weekend at Bernie’s.” And in August it argued that HP engaged in fraud by not telling Oracle that it was about to hire Léo Apotheker as its CEO and Ray Lane as its chairman when it was negotiating a settlement to a 2010 lawsuit over Oracle’s hiring of former HP CEO Mark Hurd.

But the real issue is a simple one, say people familiar with HP’s thinking in the case. Did Oracle agree to a contract with HP to continue to support Itanium — as it has been doing for years — or not?

The year 2010 was a weird one for executive moves among tech companies. Hurd resigned as CEO of HP, and took a job as president of Oracle just as Oracle was in the process of acquiring Sun Microsystems. HP sued Hurd and Oracle, and soon they settled. HP says that a clause in that settlement included a provision that Oracle would continue to port its database software to HP servers running the Itanium chip. Oracle has argued that this clause is not part of the final agreement. The settlement document itself remains confidential, but its details will likely emerge in the trial. Expect lots of arguing over different versions of the agreement.

One key part of Oracle’s argument has been that HP has been paying Intel to keep the Itanium chip alive in the face of its failure to gain traction in the mainstream server market over the last decade. This is something that HP readily concedes, since it and Intel developed the chip together in the early 1990s, and regularly renew their agreements — in 2004, 2007 and again in 2010 — to commit resources to build it and to design new generations of the chip every few years. The latest agreement calls for Intel to build two new generations of the chip.

Another argument Oracle has made concerns HP’s management during the last year. When the 2010 agreement ending the Hurd lawsuit was struck, HP was nearing the end of its search to replace Hurd. CFO Cathie Lesjak was interim CEO at the time. Ten days later, HP announced that Apotheker would become CEO. In a filing in August, Oracle argued that it never would have agreed to the Itanium partnership had it known that Apotheker, a onetime co-CEO of SAP and a figure in a contentious Oracle-SAP lawsuit, was about to become HP’s CEO. Ditto Ray Lane, a former Oracle president who was named HP’s chairman earlier this year.

People familiar with the case say that Oracle seemed unconcerned about HP’s ongoing search for a CEO, and didn’t raise any questions about it during settlement negotiations for the Hurd case. These people say that HP wasn’t deceptive, but that even if it had been it will be difficult for Oracle to argue that it’s not bound by the terms of the settlement. The language is clear and unambiguous enough that Oracle would have to argue that the Itanium clause in the agreement means nothing, these people say.

One thing is certain: The damage that HP is suffering from the ongoing uncertainty in the marketplace over its Itanium-based servers is starting to sting. HP calls these machines its “business critical” servers, and they are industrial-strength computers that aren’t sold in large numbers. Indeed, HP is the only vendor of note that even buys the Itanium chip.

But it’s a historically profitable business — HP won’t say exactly how profitable — on which HP charges its customers large service and support fees. In 2010, HP reported revenue of $2.3 billion from its business-critical operation, amounting to less than 2 percent of its $126 billion in sales that year. In 2011, HP reported that sales in its business-critical unit dropped 23 percent in the fourth quarter over the same period a year ago; sales for the year fell to just above $2 billion.

HP has argued that Oracle’s motivation is to steer HP customers toward its Sun hardware. If that was the strategy, sources briefed on the case say, it isn’t quite working out that way. One lucky party benefiting from the fight, they say, is IBM, who is winning business from some former HP customers. It’s one reason that HP is arguing for a speedy trial.

Mr. Ritchie was a longtime research scientist at Bell Labs, originally AT&T’s research division. Bell Labs announced that he died at age 70 last weekend.

Working there in the late 1960s with another programmer, Ken Thompson, Mr. Ritchie developed C while Mr. Thompson developed the original version of Unix, which is the foundation for the operating software of many mainframe computers, Web servers and Apple Macintoshes. Mr. Ritchie also made contributions to Unix.

Read the rest of this post on the original site »

Of course he used the opportunity to engage in his favorite new hobby: Taunting rival Hewlett-Packard.

The announcement was straightforward. Oracle’s forthcoming SPARC T3-4 Supercluster has achieved a performance record of more than 30 million transactions per minute, as verified by the Transaction Processing Performance Council, an industry benchmarking organization. This bested the previous record holder, a model of IBM’s Power 780 Server, which delivered more than 10 million transactions per minute.

Farther down on the TPC’s list, in fifth position, is HP’s Integrity Superdome, clocking a performance of a little more than 4 million transactions per minute. Oracle is using the results in ads that portray Oracle as a cheetah, IBM as a respectable stallion and HP as a poky turtle. (Full disclosure: This ad ran in today’s print edition of The Wall Street Journal, which like this Web site is owned by News Corp.)

At this point in his remarks, Ellison beat the war drums against HP rather loudly: “We think the HP machines are vulnerable. We think they’re expensive. We think they’re vulnerable in the marketplace. We’re going to go after them. We’re going to go after them in the marketplace with better software, better hardware, and better people.”

I’ve asked HP for a response, and will update if I get one. [Update: And here it is, courtesy of company spokesman Michael Thacker: "HP is the number one provider of enterprise servers in the marketplace. We are focused on our customers, and the combination of our technology leadership, product performance, and pricing continues to meet the needs of those customers. The numbers prove it--we experienced 25 percent revenue growth year over year during our last quarter, and we are the only major UNIX vendor that reported server growth."]

[Another update: An HP spokesperson just sent the following additional statement: "Larry Ellison bought a money-losing business that had steady market share declines for years, and which still ranks at the bottom of the market. Customers aren’t fooled by outdated benchmarks, no matter what Oracle says. HP’s market share results prove it. Sun customers are running to HP in droves because they recognize we deliver superior technology, performance and pricing."]

One thing HP and IBM have going for them is that the Oracle system in question won’t be on the market until next June.

“This set of hardware specs beat anything available to date on the tablet market,” Steven Li at Raymond James said in a research note issued today. “We believe the PlayBook tablet shows RIM is starting to compete effectively on hardware specs.”

RBC analyst Mike Abramsky offered a similar opinion, describing PlayBook’s specs as “leading-edge” and arguing that the device is well positioned for enterprise. “PlayBook may be cheaper, more productive than iPad for enterprises to deploy,” he observed, noting that the device requires no additional licenses or carrier costs and leverages existing corporate apps and infrastructure. Abramsky’s preliminarily estimate has RIM selling as many as six million PlayBooks in its first year at market.

Brian Modoff of Deutsche Bank said the company’s new OS has a lot of potential, likely making it easier for developers to write for the device and future products that use it.

Finally, over at Susquehanna, analyst Jeffrey Fidacaro talked up the device’s new QNX operating system, which sounds promising, though he noted that the transition to it could take some time, slowing the development of its app ecosystem.

“An app developer contact who works closely with RIM described the recently acquired QNX OS as “rock solid” as it was designed for mission critical applications, and felt that the SDK allowed for applications to be ported over fairly easily,” Fidacaro wrote. “The QNX OS is Unix-based and supports POSIX (easy portability of code) and OpenGL (3D graphics). We view RIM’s transition to a new OS positively as the Blackberry OS was antiquated, which apparently was not fully resolved with the recently launched Blackberry 6. However, we estimate the transition period to fully upgrade RIM’s 50+ mln subscriber base to the new platform to take at least two years. Furthermore, it is uncertain how easy RIM can port its proprietary technologies, such as security and compression, to the QNX platform.”

And that’s the real problem RIM (RIMM) is facing here: PlayBook is late to the game, and, as impressive as its specs might be, the timing of its launch– early 2011–may prove to be a real detriment. “With the PlayBook only available for consumers after the holiday season, we view this as a mild disappointment,” T. Michael Walkley of Canaccord Adams said in a note to clients Tuesday. “As such, we remain cautious in modeling RIM tablet sales due to the intensifying tablet competition and uncertain demand for this product.”

The death of the mainframe has been predicted for years, as companies opted to handle their heavy computing needs with strings of cheaper servers rather than pay a million dollars for one massive box.

The new machine, which will start shipping in September, attempts to address that challenge by letting companies use the massive computer to manage servers based on Intel Corp. or Unix technology. The software enabling the management of Unix machines will ship in the fourth quarter, and the Intel (INTC) module will come out in the first half of next year.

It took three years and a $1.5 billion in investment to produce the new machine.

Read the rest of this post on the original site

“We’re either right or we’re not. If we’re wrong, we deserve people throwing rocks at us.”

– Former SCO CEO Darl McBride in 2003

SCO’s long-running campaign against Linux may have finally been dealt a death blow. Late Thursday, the judge presiding over the company’s legal battle with Novell (NOVL) rejected its request for a new trial and upheld an April jury decision that determined Novell, not SCO, is the rightful owner of key Unix copyrights.

“SCO argues that it is entitled to judgment as a matter of law because the ‘verdict cannot be squared with the overwhelming evidence and the law,’” Judge Ted Stewart wrote in his decision. “The Court respectfully disagrees. The jury found Novell’s version of facts to be more persuasive. This conclusion is well supported by the evidence. There was substantial evidence that Novell made an intentional decision to retain ownership of the copyrights….Therefore, SCO is not entitled to a new trial.”

So it seems that after seven years, this charade by SCO is finally over. What will it tell the bankruptcy court now? Earlier this year, former U.S. District Judge Edward Cahn, the company’s bankruptcy trustee, told the Salt Lake Tribune that the jury decision will not dissuade it from pursuing its lawsuit against IBM (IBM). Said Cahn: “The copyright claims are gone, but we have other claims based on contracts.”

Thursday’s ruling would seem to be a setback for those plans, but SCO has a history of pressing on in the face of repeated defeats. We’ll have to wait to see if one of its Iraqi Information Minister-style press releases is forthcoming.

In any event, the stoning has begun on Wall Street, where SCO stock is trading around five cents a share.