Thanks for the mention of Kickfire. On the question of Kickfire vs Moore’s Law I would point your readers’ to a blog posting we did on the topic: http://www.kickfire.com/blog/?p=19. In essence, the Kickfire chip architecture is based on dataflow, not instruction-flow. The dataflow architecture is fundamentally more efficient for large data volume processing. This is why, as has happened in other industries before, once the transition to dataflow has been made it is next to impossible for general-purpose CPUs to catch up. This is why Nvidia still leads in graphics, not Intel, and why Juniper and Cisco use specialized routing chips not processors from Intel or AMD.

Btw, Oracle’s Exadata server is much more akin to Netezza than to Kickfire. Both use hardware to filter from disk to memory. In both cases though SQL is still executed on general-purpose CPUs, not on a SQL chip.

Karl

thanks for passing on the news, much appreciated.

Just wondering about this:

“By the way, I made a bit with Paul Cunningham that Kickfire would fail because of Moore’s law.”

Maybe it’s ignorance on my part, but I don’t see what you mean. Genuinly interested, can you pls. explain? TIA,

Roland.

By the way, I made a bit with Paul Cunningham that Kickfire would fail because of Moore’s law.