On openness, integration, and the importance of great software
The topic of platform openness vs. tight hardware integration, and the importance of great software is once again lighting up the blogsphere, thanks to comments made last week by Apple’s CEO, Steve Jobs. The iPhone and iPad maker’s remarks, directed towards rival Research in Motion, were biting. He accused RIM of merely being a hardware manufacturer with little concern for software. He accused most of his competitors of throwing hardware at a problem and assuming the same old software would suddenly, magically come alive to support it. It won’t, of course.
The same is true in our market. No matter how much appliance vendors have tried, taking tired transactional RDBMS software and adding boxes or blades increases costs more than it increases performance. Certainly, with 128 CPUs, hardware can make bad software better (as at least one vendor demonstrated last week), but if you open the market to competition, then you force vendors to optimize their software. To be the best that they can be.
Teradata might need 128 CPUs focused on one task to take your data warehouse to warp factor 9. Great software, however, performs no matter how many CPUs you throw at it, and adjusts with elastic scalability. That’s its job.
Like Apple’s Steve Jobs, we think that focusing on great software and on what we do best is the winning formula both for us and for our customers.