Oracle Exadata

Off topic, where I brag a little about my job

I don’t work directly on Oracle Exadata, but I am responsible for some of the underlying software for the compute part of this thing that we call Exadata.



Reducing power, cooling and space requirements is something all first tier server companies have been working on for a few years now, though we tend to do a little bit better job with the most dense servers on the market. Beyond that, Oracle’s record breaking I/O capabilities is kind of a Big Thing. There’s an orders of magnitude difference in transaction performance between Exadata and most other machines out there, and we do it in a smaller machine with significantly less power and for less money.

Exadata isn’t something for your home computing needs — these are multi-million dollar machines with a couple of hundred CPU cores, terabytes of RAM for data warehousing and online transaction processing where we push the envelope in I/O and compute technology.

The state of practice in high performance computing involves clustering hundreds to thousands of discrete computers, which requires very very fast I/O (usually InfiniBand) connected via fast, massively interconnected switches. We combine this fast I/O with caching on solid state “disk” storage to accelerate database performance to levels that were unheard of even a year ago. I like that I’m peripherally involved in the design of some pretty slick stuff like this.

More –> Oracle Exadata Database Machine.

3 Comments

  1. We love it when you talk Geek, Fritz! Now, ‘splain to us how them internets tube things work.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.