Tuesday, April 28, 2009

SAP shores up SAP NetWeaver BW with Teradata

NetWeaver Business Warehouse (BW) will run on Teradata databases, a move that analysts say is great news for SAP NetWeaver BW customers.

The move shores up SAP NetWeaver BW's deficiencies, which include a reputation for slow query responses, limited data volume scalability, and limited visibility into non-SAP data, according to Philip Russom, senior manager of research at The Data Warehousing Institute (TDWI).

"The new SAP-Teradata combination has great potential for curing these weaknesses," Russom said in an email response. "Thanks to the SAP-Teradata combination, I feel we'll see even more BW implementations in the future."

The Teradata option will appeal to customers that have Teradata already and were forced until now to run SAP NetWeaver BW on a different platform, according to Boris Evelson, principal analyst at Cambridge, Mass.-based Forrester Research. NetWeaver BW is currently supported only on Oracle, IBM and Microsoft SQL databases. It will also appeal to SAP customers growing through mergers and acquisitions that will need to accommodate larger sets of data, according to Evelson.

But don't expect SAP customers who are running NetWeaver BW on SQL Server to rush to move to Teradata, he said. Teradata is still a very expensive proposition compared with SQL Server.

Before this partnership, if a company ran SAP and Teradata, the data was stored as relational data in Teradata, according to Mark Whitehorn of PenguinSoft Consulting Ltd. A copy -- called a shadow copy -- was extracted and held in SAP. It was then restructured, and SAP analytics ran against that restructured shadow copy, Whitehorn said in an email response.

In the new model, Teradata is the data warehouse layer, above that is SAP NetWeaver BW and NetWeaver BW Accelerator, and above that, the front-end analytical tools, according to Whitehorn. The data is held in Teradata as before, but no shadow copy is produced. So there is only one data warehouse, which is held as relational data in Teradata.

The move enhances the back end of SAP's business intelligence portfolio. While SAP's acquisition of Business Objects gave it a great front end, the back end was still lacking, Evelson said. The data warehouse in SAP NetWeaver BW is more of a departmental data mart, he said -- it never scaled to hundreds of terabytes, it didn't have lots of the infrastructure an enterprise-class data warehouse should have, and it was never very stable.

"If you look at the SAP business intelligence portfolio, it is very heavy on the front end as opposed to the back end," Evelson said. "Now, basing BW on top of Teradata, that's great."

Teradata provides better storage and scalability in the 25 terabyte area, according to Tim Lang, vice president of product management for SAP BusinessObjects.

The new partnership also enhances support for Teradata, Lang said. SAP customers can call their SAP support line if they have issues with Teradata.

"There's a much stronger connection on the support side as well," he said. "There's a lot of advantages we bring together in a joint solution, particularly when talking about very large datasets."

Russom thinks the SAP-Teradata partnership will spur larger SAP BW implementations because of the scalability assured via Teradata. Likewise, improvements to query performance and broader access to enterprise data made possible by this combination will lead to more SAP ERP users adopting "operational business intelligence" and other time-sensitive, data-driven business practices, he said.

The partnership also heightens the speculation that SAP is interested in acquiring Teradata -- a rumor that surfaced earlier this year, according to Madan Sheina, principal analyst with London-based Ovum Research.

The acquisition would be a wise move on SAP's part, Evelson believes. As consolidation in the industry continues to lead the news of the day (e.g., Oracle's recent Sun acquisition ), partnering with companies will become increasingly difficult.

"I think [buying Teradata] would make a lot of sense," Evelson said. "The major piece SAP is missing is really an enterprise-class data warehouse."

But Sheina said that the Teradata-SAP partnership also underscores something that is becoming a competitive advantage for SAP -- its continued emphasis on platform neutrality.

"SAP is perhaps unique among the big four BI plays in that it is agnostic about which database platform to do data warehousing on," Sheina said in an email response. "Such an option also gives SAP an opportunity to earn revenue for itself rather than [its] biggest applications [and BI] rival."

Monday, April 27, 2009

Let the IT Land Grab Begin

With Oracle ready to annex Sun Microsystems, there may finally exist a true, full and global counterweight to IBM's hegemony. Who else loses? Microsoft, SAP and Cisco -- and maybe even Amazon. If HP plays along and aligns with Sun and Oracle, together they could create a full-service IT powerhouse.

The reported acquisition of Sun Microsystems (Nasdaq: JAVA) More about Sun Microsystems by Oracle (Nasdaq: ORCL) More about Oracle makes a ton more sense than IBM's (NYSE: IBM) More about IBM earlier failed bid. This new compact, if it succeeds, will bring as good an end to an independent Sun as the pioneering (yet long flagging) IT vendor could have hoped for at this sorry stage in its history.

However, there are much larger implications in Oracle's latest super-grab than Sun's demise and assimilation. Among them is the fact that IBM now -- for the first time, really -- has a true, full and global counterweight to its role and influence. Oracle plus Sun aligned with HP (NYSE: HPQ) More about Hewlett-Packard (which I fully expect) meets and begins to beat IBM at all the important full-service IT games.

This is truly healthy for IT and the global IT marketplace. IBM's earlier purported bid for Sun always smelled bad to me. It was, it turned out, mostly a red herring. Perhaps Oracle needed the IBM roller coaster ride to focus its intentions. Nonetheless, the outcome is optimal. It bodes well for cloud computing too, as Oracle just about overnight becomes a cloud force to reckon with. I always thought Larry Ellison was just biding time on this one. The recession has hastened the timetable.
Hello, Losers

Other than IBM's unassailed hegemony, the other losers in this are Microsoft (Nasdaq: MSFT) More about Microsoft (actually possibly creeping to irrelevancy faster than anyone could have imagined three years ago), SAP (NYSE: SAP) More about SAP AG and Cisco Systems (Nasdaq: CSCO) More about Cisco Systems. Amazon (Nasdaq: AMZN) More about Amazon.com may also be getting more competition soon on the Platform as a Service front. Using Sun's cloud investments, implementations and plans, Ellison can also quickly forge together his own counterweight to Salesforce.com (NYSE: CRM) More about Salesforce.com. No need to buy it now (for a while).

Open source in general, too, may take a hit, as I don't expect Unbreakable Linux to remain Oracle's point on the operating system arrow. Solaris will be the prime Oracle OS for performance, meaning Oracle's channel pipeline to Red Hat (NYSE: RHT) More about Red Hat will shrink. MySQL will be a means and not an end for Oracle, which would, of course, prefer an Oracle 11g cloud.

Suffice to say that whatever momentum Sun had behind open source More about open source everywhere will be muted to open source sometimes as a ramp to other Oracle stuff, or to grow the community and keep developers happy.

Like IBM, Oracle will have little interest or need for open source middleware or service-oriented architecture (SOA) components. Further, given Oracle's early and deep interest in Eclipse and OSGi, the Java More about Java tools will stay free and open (with a lot of Oracle wizards embedded across the database and other middleware). The tussle for influence between Oracle and IBM in Eclipse and the Java Community Process (JCP) will be great fun to watch in coming years. Again, this is healthy. (Good thing Sun opened this up, eh?)

No other company has shown an ability to merge and integrate at the massive scale and complexity that Oracle has. Its acquisition spree that began five years ago is unprecedented in its scope and level of success. We have no reason to suspect that the way it handles Sun will be any different.
As for the Winners

Winners on the deal include Java itself in the fullest and broadest sense. Oracle and IBM are the premier Java vendors, and the might of IBM (and its customers and developers) in the market will force Oracle to keep Java open and vibrant, while Oracle's penchant for control and commercial success will keep Java safe and singular. I expect the old BEA (Nasdaq: BEAS) More about BEA Systems WebLogic implementations now at Oracle to gather some minor bundles from Sun's software portfolio, but Sun's enterprise software stack (for all intents and purposes) is history. I can't see Glass Fish or Net Beans going anywhere but bye-bye. Same with the Sun SOA stuff.

Most interesting will be the way that Oracle matches the Sun assets against HP's burgeoning partnership with Oracle. Will HP perhaps buy Sun's hardware, storage and integrated circuits intellectual property outright after the Oracle acquisition is final? I'd bet on it.

The Exadata announcement last fall is a good example of what to expect. Business intelligence is the killer enterprise application of the day (era), and Oracle and HP aim to win. Coupling Oracle BI and business applications is something special ... better potentially than what IBM and SAP can do. Should we expect from this Oracle-Sun merger some more love or more between IBM and SAP. Oh, yeah!

We should expect to see a major go-to-market push by HP and Oracle, with all kinds of appliances and solutions portfolios. Both Oracle's and HP's love of virtualization allow all kinds of neat packaging. Expect some of the industry's premier on-premises cloud solutions ASAP.

Indeed, we now have a land grab race for the modernized data center/private cloud between Oracle/Sun/HP and IBM. What's more, HP with all the old DEC stuff, plus Sun's Unix, may keep Unix alive and well while keeping IBM at bay with its everything mainframe lust.

On the blue sky front, consider if Apple (Nasdaq: AAPL) More about Apple and Google (Nasdaq: GOOG) More about Google get closer to the Oracle-Sun-HP trifecta? Wow. Cloud city.

Larry Ellison correctly predicted a few years ago that only a few IT companies would remain. Maybe we should just remove the "IT" and keep it at only a few companies will remain -- and Oracle will be one of them.

Talk about pure irony. It was when Oracle turned its back on Sun four years ago with the Unbreakable Linux and Java process business (Eclipse over NetBeans, OSGi support, etc.) that Sun's nosedive deepened. In a sense, you could say that Oracle pushed Sun off a cliff in slow motion, only to catch the pieces at fire sale prices.

Thursday, April 23, 2009

Microsoft sales fall 6 percent from a year ago

As analysts predicted it might, Microsoft on Thursday reported the company's first ever year-over-year sales decline for the quarter ended March 31.

The software maker said fiscal third-quarter sales totaled $13.65 billion, down 6 percent compared with $14.45 billion in the same quarter a year ago. Its per-share earnings were 33 cents per share, although that included severance and investment impairment charges that reduced earnings by 6 cents per share.

Analysts had been projecting sales of $14.15 billion and per-share earnings of 39 cents, down from 47 cents a year ago, according to Reuters Estimates.

Microsoft had said in January that the crystal ball for the company was cloudy and at the time announced its first companywide layoffs, with plans to chop 5,000 jobs over an 18-month period.

"While market conditions remained weak during the quarter, I was pleased with the organization's ability to offset revenue pressures with the swift implementation of cost-savings initiatives," Microsoft Chief Financial Officer Chris Liddell said in a statement.

The company noted that software sales to large businesses were stable during the quarter, but that weakness in server and PC sales hit its Windows, server and Office units.

Whereas Intel and EMC have been somewhat optimistic that things may have reached bottom last quarter, Microsoft's comments were less hopeful.

"We expect the weakness to continue through at least the next quarter," Liddell said.

The company didn't have much to say on several closely watched topics. The company did not give a specific sales or earnings outlook for the coming quarter, instead only noting what it expects as far as its operating expenses.

As for Windows 7, Microsoft just noted that it "remains on track for a fiscal year 2010 launch." That's even less specific than its usual comment, which is that it should ship within three years from general availability of Windows Vista, meaning by January. The software maker has been pushing to have Windows 7 out in time to be on PCs by this year's holiday season, with recent indications that the company is still aiming for that goal.

Shares closed Thursday at $18.92, up 14 cents. In after-hours trading, investors sent Microsoft shares higher. The stock was trading recently at $19.85, up 93 cents, or nearly 5 percent.

The PowerPoint slides that Microsoft put out to accompany its earnings report offered a few more nuggets. The company saw its online advertising revenue decline 16 percent, causing that unit to fall below what analysts were expecting.

PC unit sales were down 7 percent to 9 percent during the quarter, but the industry's revenue dropped more than that as Netbooks continued to make up a larger slice of sales--a trend that hurts both the PC makers and Microsoft. Microsoft sold 1.7 million Xbox 360s during the quarter, up 30 percent from a year ago and helping push that unit back into the red.

Here's a look at how each of Microsoft's individual units did during the quarter, in terms of both revenue and operating income.

Source:Cnet News

Tuesday, April 21, 2009

Voracious Oracle Gobbles Up Sun

Oracle's corporate org chart is starting to look like one of those maps of the Roman Empire or of Alexander the Great's conquests or something. It already covers great swaths of the technology industry, and it just keeps spreading in every direction.

Consider for a second this paragraph, courtesy of Bloomberg:

"Oracle is entering new markets as it seeks to reach $50 billion in revenue by 2012. The purchase is Oracle's third-largest after its $10.3 billion takeover of PeopleSoft Inc. in 2005 and the $8.5 billion purchase of BEA Systems Inc. last year. Oracle has spent almost $34.5 billion on purchases since 2005 to buy 52 companies, making it the most acquisitive software company in the world."

"The purchase" mentioned in that paragraph, of course, is Oracle's planned acquisition of Sun Microsystems, revealed yesterday. Oracle, once mostly a database vendor, is all over the enterprise technology map now, so to speak. Enterprise applications (ERP and CRM), business intelligence, storage, database technologies, severs -- you name it, and Oracle's probably trying to sell it to corporate customers, or will be soon enough.

Yes, the Sun purchase would make Oracle a hardware company by delivering to Larry Ellison's empire Sun's server lines, which are already home to lots of Oracle databases in enterprises worldwide. (Intriguingly, there is some speculation that Oracle didn't want the hardware business and won't keep it. The talk is that Oracle just wanted Sun's software business but had to take the hardware as part of the deal. For now, though, Oracle appears to be soaking in the entirety of Sun's operations, and Ellison is talking about keeping everything for himself.)

The acquisition announcement had jaws dropping all over the industry. Steve Ballmer said he was "very surprised" and needed to think about the whole thing, a rare moment in which the Microsoft CEO seems to have been caught off guard. Some analysts expressed shock, while others took an approach that might be described, if we want to be cliché (and we do), as "cautiously optimistic."

For Microsoft partners, it's hard to say exactly how an Oracle-Sun marriage will play out. HP and IBM will more likely be concerned by the massive shift in the competitive landscape at the outset of the deal than will Redmond.

But a stronger Oracle with a more diverse product offering (to say the least) could speed the spread of Linux-based servers in the enterprise, given that Oracle as a company is kind of a Linux fan and definitely not so much a fan of Microsoft. Of course, buying Sun would make Oracle a major factor in the Unix operating system game. Everything considered, an Oracle-Sun combo probably doesn't offer many positives for Microsoft partners, unless the two companies stumble over each other in the integration process.

Which, if history is any indicator, they probably won't. Ellison and his troops have been amazingly adept at swallowing companies and making them productive parts of the Oracle empire. Despite an economic downturn and loads of purchases in recent years, Oracle has just kept rolling along earnings-wise, beating the Street again with its latest quarterly report.

Oracle-Sun is another mega-deal that signals that the age of city-states is long over in the technology industry, and the age of empires has well set in. The thing about empires is that they usually end up collapsing under their own weight (right, Wall Street?). Thus far, Oracle has avoided that fate. It probably will with the Sun buyout, too. The question now is: Where will Larry Ellison go next?

Source:rcpmag.com