Intelligent Enterprise

Perhaps I've been watching too much political coverage on TV lately, but at one point during IBM's Information On Demand (IOD) press conference yesterday, it struck me like a campaign stop. To set the scene, the first question during the post-announcement Q&A session came from Forrester Research analyst Jim Kobielus, who cited the recent "pretty significant" announcements by Oracle and Teradata in the area of data warehousing and appliances. Noting the lack of warehouse- and appliance-related announcements at IOD, Koblielus asked, "what is IBM's strategy, going forward, to make your InfoSphere Balanced Warehouse portfolio ever most cost-effective and ever more scalable?"

Rather than responding directly to the question, Arvind Krishna, Vice President Data Management and Worldwide Information Management Development, first took on the role of Oracle attack dog. "The only pricing Oracle has provided [on the Oracle Database Machine and Exadata Storage Server] is for the hardware," Krishna challenged. "When you actually add in the pricing of the software, it's significantly higher than our pricing — more than three times as much. If you add in the price of the software to the $40,000 per terabyte that they claimed, the price is closer to $300,000 per terabyte."

Wow, that sounds high, so let's call in the Intelligent Enterprise "Truth Squad." According to Curt Monash, our best expert on this topic and someone who has gone over these announcements with a fine-toothed comb, the price per terabyte of the Database Machine will actually come in somewhere between $60,000 and $198,000 per terabyte, including software, depending on the drive and processor configuration.

Krishna then challenged the Oracle Database Machine on performance, saying "it's actually far less than the performance of the Balanced Configuration units we've already offered… "

Truth Squad: Hmmm… we'd love to see the real-world testing behind that claim. Even impartial, third-party analysts haven't seen Oracle's hardware in action, so we have yet to see real, objective performance measures.

Krishna finally responded to the real question about what IBM would offer by saying, "that does not mean that we're sitting still or that we're not looking at not only significant incremental improvement, but innovation that will lead to us getting a five-X to ten-X improvement within the next few months." The vague "that doesn't mean we're not looking at…" construction sounded like political double talk, but I'll hold them to the campaign promise of a "five-X to ten-X improvement within the next few months."

As a campaign watcher, I'd have to say that of the three database incumbents — Oracle, IBM and Microsoft — Oracle seems the closest to delivering real appliance change to the market. IBM is promising something new in appliances in "the next few months" while Microsoft has said its appliance-ready database won't be available until the first half of 2010. If you want to elect a known quantity, I have to give this year's election season to the independents.

E-MAIL  |