>Then Intel a couple of years later rolled out Nehalem, with an integrated memory controller and hyperthreading, cementing their advantage in the server market. AMD has been playing catch up ever since.
Before intel essentially copied HT from AMD with their QPI, (I believe that Nehalems were the first QPI xeons) AMD servers nearly always came in dramatically lower power than the FBDIMM using Xeon-based servers.
Also note, in the early days of hyperthreading, it was a great way to run your two active processes on one core, while your second was idle. My understanding is that even now, in the best case, it's not a particularly huge boost.
I mean,yeah; between the release of the QPI xeons and now, for most things, intel has had the superior chip. But before QPI? man, if you paid for your own power, AMD was dramatically superior for high-ram applications.
Hyperthreading is better now. They even had to add another instruction to fix some of the issues. Plus a lot of work on the scheduler. Not worth disabling now!
Before intel essentially copied HT from AMD with their QPI, (I believe that Nehalems were the first QPI xeons) AMD servers nearly always came in dramatically lower power than the FBDIMM using Xeon-based servers.
Also note, in the early days of hyperthreading, it was a great way to run your two active processes on one core, while your second was idle. My understanding is that even now, in the best case, it's not a particularly huge boost.
I mean,yeah; between the release of the QPI xeons and now, for most things, intel has had the superior chip. But before QPI? man, if you paid for your own power, AMD was dramatically superior for high-ram applications.