StealthX32
Senior member
Alrighty. I currently have an AMD Athlon T-Bird 800, and I haven't really spent much time lookin' into computers since the Celery 300A craze. So I'm kinda outta the loop. A couple questions before we start. Back in the good ole days, 4.5x100 meant 100 MHz FSB, which translated into a 100 MHz memory bus. Now, back to today. According to this picture (AMD System Architecture - stolen off HardOCP), the memory bus and the FSB run completely separately at separate speeds, right? So this means, I don't have to have a 200 MHz FSB to run DDR400 at full speed, right? Now, I see some people running their processors at 200 MHz FSB...such as TonyB. What would the advantage be running a 200 MHz FSB? I'm assuming you'd unlock the multiplier and drop it down to allow the significantly higher FSB, but since you don't need to match FSB and memory bus speed, why is this necessary?
So, now, assuming I've got the AMD architecture correctly, and that I want to spring for an Asus A7N8X, and want to get Corsair XMS PC3200 memory...would I be better off with a 2200+ T-Bred A or pay up an extra 20 bux for a 2400+ T-Bred B? Or a 2200+ T-Bred B? I'm assuming I can get the XMS PC3200 to run at 400 MHz (since Anandtech says its possible 🙂 ). And reading a couple articles off Overclockers.com, I came across one article that tells me how to tell processors apart from others, and another helpful article is one from Anandtech that illustrates the differences between the A and B cores. From what I gather, the T-Bred would be easier to overclock, but other than that, there shouldn't be a difference.
So, my primary concerns are probably not overclocking but rather system stability. From what I hear from the nForce2 reviews, a T-Bred B's multiplier would be unlocked without any additional modifications. Which would be nice, since I'd be able to tinker with the FSB without any additional work...and assuming I'll get performance gains from it. So, what do you all think? I was gonna get this all from Newegg.com since shipping would be free, and they have everything in stock, but they only have the 2200+ in a T-Bred A core (according to that article from Overclockers.com). So 2400+? Grah. Straighten me out please. Thanks all!
So, now, assuming I've got the AMD architecture correctly, and that I want to spring for an Asus A7N8X, and want to get Corsair XMS PC3200 memory...would I be better off with a 2200+ T-Bred A or pay up an extra 20 bux for a 2400+ T-Bred B? Or a 2200+ T-Bred B? I'm assuming I can get the XMS PC3200 to run at 400 MHz (since Anandtech says its possible 🙂 ). And reading a couple articles off Overclockers.com, I came across one article that tells me how to tell processors apart from others, and another helpful article is one from Anandtech that illustrates the differences between the A and B cores. From what I gather, the T-Bred would be easier to overclock, but other than that, there shouldn't be a difference.
So, my primary concerns are probably not overclocking but rather system stability. From what I hear from the nForce2 reviews, a T-Bred B's multiplier would be unlocked without any additional modifications. Which would be nice, since I'd be able to tinker with the FSB without any additional work...and assuming I'll get performance gains from it. So, what do you all think? I was gonna get this all from Newegg.com since shipping would be free, and they have everything in stock, but they only have the 2200+ in a T-Bred A core (according to that article from Overclockers.com). So 2400+? Grah. Straighten me out please. Thanks all!