Hi, Im planning on purchasing a new processor and mobo within the next few months. an AMD XP 2700+ and a A7N8X-X ASUS motherboard. I'm wondering if the addition of 700 MGHZ and a faster mobo will increase performance over my current 1.4ghz AMD thunderbird??? How come AMD processors have much less cache and FSB???? does this mean they perform worse, like comparing a P4 2.1ghz to a XP 2700+, would the XP be slower even though it has the same Gigahertz? I heard a while ago that AMD builds there processors better, so Intel has to make up for it by putting more FSB and Cache, is this true? Please dont be bias, it doesnt help when Die hard (company name here) fans respond with one-sided responces. Thanks ahead of time!

"I heard a while ago that AMD builds there processors better, so Intel has to make up for it by putting more FSB and Cache, is this true?" No. That is ignorance, even if you design microarchitectures. First off FSB and cache are very important and you want to have as a fast a FSB and as much on-die cache as you can. So making up for anything with solid, physical resources is not bad design. "How come AMD processors have much less cache and FSB????" Hard to answer. Design timelines are difficult for consumers to see, but basically AMD was ahead with their Athlons vs. P3's and early P4's, but has lost that top ground (arguably) to Intel for this moment. The design timeline is probably focused heavily on the Opteron. It is hard to stay on top of the curve constantly. Mostly, you try to beat it at points where you can grab market share and ride that even as you slip down the curve. But there is far more to this than simply AMD and its chips (the market, device manufacturers, silicon foundaries, etc.). If you really wanted a better designed CPU, I would look into a PowerPC 970, otherwise known as a G5, designed by IBM for Apple's new line of 64-bit computers.

AMD's used to be slower in a side by side performance benchmark when compairing like for like. For example a 1.0 GHz AMD was considerably slower than a 1.0GHz Pentium. Now however they are virtually identical with a slight (VERY slight) edge going to the AMD. MAC's are not a viable option in my opinion. They are sideline players and always will be until they get with the program and go mainstream. I have no dislike for them but they are still the VW bug of PC's. Unless you edit movies or graphics for a living they are just simply not software and/or hardware diversified enough to be a contender. Jimi_l

thanks for the help, but p4s are way overpriced, are they TRYING to be Microsoft?! :|

"I heard a while ago that AMD builds there processors better" put complex in place of better and you have it right. That's why the xp 1500+/2000+/2400+ cpus are equal to the p4 1.5g/2.0g/2.4g cpus. AMD cpus do more work per clock cycle. "so Intel has to make up for it by putting more FSB and Cache, is this true?" Intel does not use more cache and fsb as a way of making up for some deficiency, its just to make their machines as fast as they can. It is true that Intel cpus require more cache and fsb speed in order for the cpu to be optimized. If the xp 3000+ and 3200+ did not exist, the world would be simpler. You could just believe the amd rating numbers (like 2400+) were equal to their P4 equivalent (like 2.4g). But because amd overrated the 3000+ and 3200+, we have introduced confusion. "like comparing a P4 2.1ghz to a XP 2700+, would the XP be slower even though it has the same Gigahertz" The p4 2.1g is about as fast as an xp 2100+. If you want gauge fsb effect, 1) assume a 533fsb p4 is equal to a 266fsb xp, 2) and that an x% increase to fsb alone (leaving cpu mhz constant) increases the system performance by 0.25X%. That is, a 2.4g/800fsb is 12.5% faster than a 2.4g/533fsb. Note that all current xp cpus can be set to whatever fsb the motherboard can handle. If you have a 2400+ the default is 266fsb (15x133 = 2.0g). But if your motherboard can handle it, you are much better off running it at 444mhz fsb (9x222 = 2.0g). This increases the system performance by 0.25 x 222/133 = 17% without overclocking the cpu.

oops. Wrong number. 0.25 should be 0.40 "2) and that an x% increase to fsb alone (leaving cpu mhz constant) increases the system performance by 0.25X%.... This increases the system performance by 0.25 x 222/133 = 17% without overclocking the cpu." actually an x% fsb increase produces an 0.4x% system performance increase. So the example should say ... "this increases the system performance by 0.40 x 222/133 = 27% without overclocking the cpu."

More cache is not necessarily better. Larger caches take longer to search for required data. So there is normally an optimum size for a cache to be.