Well silly me.
I guess I just get the impression that DDR is twice as SDR based on what we've seen with video card memory.
Why wouldn't a FSB being DDR'd or QDR'd be the same way????
And how did Tom test that the 100MHz DDR for the Athlon was only 7% better than an SDR version? I hope that score doesn't factor in other bottlenecks into the equation, because if it does then we wouldn't know for sure (for example, the P3 1 GHz _is_ twice as fast as the P3 500 MHz, but a system like that isn't going to be 100% faster).
Did Tom adjust the score to other bottlenecks? I would think that if there was something that only used the FSB, it would be twice as fast on a 100MHz DDR FSB than a 100MHz SDR bus.
In fact, you talk about Tom's review of the DDR mobo, in which case you are comparing the MEMORY, not the FSB. The Athlon has _ALWAYS_ had the 100MHz DDR FSB (even while using 100/133 MHz SDR memory).
Now that I look over your message, you were talking completely about the Athlon DDR MOTHERBOARD (which I was not).
BTW, in applications where memory is a huge bottleneck, a DDR RAM system will perform pretty much twice as fast, barring any other bottlenecks of course.
In fact, the more and more I read your argument, the more I see you didn't understand what I had written. I had said a 100MHz DDR FSB is 200 MHz effective. I did NOT say it would give you 2X the performance. Just because something is running at twice the speed, doesn't mean it is running twice as fast, and I never once said that it did.