One, in terms of playing the game, and can not simply say that which CPU is good, in the same level of CPU, their performance is almost the same, but AMD's CPU to take the algorithm is different, so their performance is also a little bit different

Currently, the difference between the INTEL and AMD's CPUs, and due to the difference between the difference brought Performance and efficiency differences have the following brief points, for reference only:

1, from the single-crystal silicon process: INTEL: 0.09 (to reduce the cost, increase the number of transistors), AMD: 0.13 (the cost of higher than 0.09), which leads to the same proportion of the price in the are lowered, INTEL or earn money, and AMD will not earn too much money at the very least ah, or even also accompanied by money (loss), although the market share has increased, all of which led to the recent AMD lawsuit

2, from the pipeline: INTEL: 31 levels (can be raised to a higher main frequency, but brings greater heat: for example, the P4-670 super to 7.4G, but have to use liquid nitrogen to dissipate the heat, but also prone to cause inefficient execution of the instructions, so to get out of a hyper-threading); AMD: 20 levels (can be raised to higher main frequency, but brings greater heat: for example, the P4-670 super to 7.4G, but have to use liquid nitrogen to dissipate the heat, and easy to cause instruction execution, so get a hyper-threaded (to make up for); AMD: 20 level (the efficiency of instruction execution is stronger than the 31 level, but the frequency is limited to improve and relatively low heat, efficiency and frequency are two different directions of development, mainly depends on the user's choice)

3, cache: INTEL: 1 level 16K, 2 level 1M-2M (integer arithmetic as well as gaming performance is not as fast as AMD's (and there is a major reason for this), but it's a good idea to use a super-threading system to compensate for this. (There is a major reason for the role, later on), but for the network and multimedia (floating-point computing) applications than rivals strong

AMD: 128K level 1, level 2: 512K (integer computing fast, game performance is good, but for multimedia applications a little inferior)

4, the memory management framework: INTEL's memory management framework is still the use of the traditional way to manage the North and South Bridges of the motherboard (causing the CPU and memory management). Management (resulting in a large delay in data transmission between the CPU and memory, for the implementation of the game is not as good as AMD's, but for future upgrades to reduce costs) AMD is the CPU internal integrated memory controller (reduces the delay in data transmission between the CPU and the memory (for the enhancement of the performance of the game has a considerable role, but also the main reason for what was said, but also to make up for the 2 level of only 512K).

5. Instruction set INTEL: MMX, SSE, SSE2, SSE3, EM64T

(Most of the games and software are based on INTEL's instructions, which are optimized for INTEL. Optimized for INTEL, but the 64-bit instructions have compatibility drawbacks for the new 64-bit systems now, so recently they have to be compatible with AMD's X86-64 instructions, and the CPU stepping value has been changed from E0 to G1.) AMD: 3DNOW+, MMX, SSE, SSE2, SSE3, X86-64 (There are 2 instructions missing from the supported SSE3, but it's not a big problem because those 2 are specially designed for INTEL HyperSpeed, which is the only one that can be used in a 64-bit system. But it's not a big problem, because those 2 instructions are specifically for INTEL's Hyper-Threading technology, and it's fine if they're not there, AMD doesn't support Hyper-Threading technology anyway. Due to the fact that AMD's 64-bit technology originated from DEC's Alpha technology (one of the 64-bit technologies), coupled with AMD's own 2 developments, it led to the rapid emergence of 64-bit technology in the civilian market, and the 64-bit system of Microsoft had to be based on AMD's X86-64 development (who asked AMD to do so?). 64-bit development (who called AMD first launched the civilian 64-bit it), in order to eliminate as soon as possible for the 64-bit WINDOWS compatibility issues, INTEL was also forced to start compatibility with AMD's 64-bit instructions (not that INTEL does not have the technology to develop 64-bit, but because of its market strategy led to its very passive, missed the best time to launch the 64-bit, so that the AMD 64-bit on the stand! The 2 companies still have to look at Microsoft's face in the end, from this point of view, they haven't fully reached the market monopoly position --- hardware manufacturers still have to look at the face of the software giants, really sad!)

To summarize: now who is more cost-effective is to look at the user's scope of application (and inevitably by the scope of the application to decide), and not simply by the price to decide, I do not agree with the so-called poor people only use AMD's statement. As INTEL feel the pressure from AMD so the company in the development strategy to make a major decision to change (from the pursuit of frequency to the pursuit of performance changes, but also had to give up by INTEL's own creation of Moore's Law this myth, the full shift to the improvement of the performance of the CPU, the CPU in the frequency of 3.8G drew a small period, so that the goal of 10G to become a bubble; with reliable sources of information. With reliable news: INTEL's future CPU frame enough will be based on the technology of the current mobile CPU, and put forward the concept of sex consumption ratio (rather than cost-effective) and has recently successfully developed samples, as far as performance is concerned, will be 3 times the current P4 - 5 times, and power consumption from the laptop CPU's 5W to the desktop CPU's 35W to the server CPU's 65W, the core will be a dual-core or 4-core CPU, the core will be a dual-core or 4-core CPU. Will be dual-core or 4-core, front-side bus: 533MHZ, 667MHZ, 800MHZ, 1066MHZ; no more hyper-threading technology (because there is no need, the emergence of hyper-threading technology is mainly to make up for the inefficiency due to the excessive length of the pipeline, the new INTEL's CPU will not be used again with the 31-stage pipeline, there may only be less than 20 or less) The new INTEL CPUs will not use the 31-stage pipeline anymore, probably less than 20 stages or lower, and the frequency will not exceed the existing frequency (which means that 3.8G will be the highest frequency of INTEL now and even in the future). In the coming 2007, the big battle (INTEL and AMD) will be a showdown, and it will be clear who will win and who will lose, and who will be good and who will be bad, and whose price/performance ratio or sex/hour ratio will be higher, and to be honest, it is a little bit worrying for AMD (AMD has said that it will not change the existing CPU frame). To be honest, I'm a bit worried about AMD (AMD has recently said that it will not change its existing CPU architecture), but I still believe that time will tell! Who will be the product that consumers should look forward to the most? I'm sure it will be in the near future!

For AMD, its most popular aspect is its good overclocking performance and low price, which is the fundamental reason why it currently holds the market share of processors and its advantage. In our choice, if it is a DIY master, then choose AMD is certainly right, can spend less money to get better performance, the price of the same main frequency AMD and Intel, the former price is only about half of the latter, and now most of the AMD processor motherboards have fools overclocking software, although it can not be overclocking to play to the limit, but also can overclocking addiction. AMD's heat problem has always been the most concerned about the problem, in fact, not, now AMD's processor more added overheating protection chip, so the heat problem has basically been solved, do not have to worry about.

When buying AMD products, it is important to note that due to its good overclocking performance, some * merchants have started to sell low-frequency versions of overclocked and then polished products, how to identify whether the products are polished or not, the easiest way is to look at the processor's L2 and L3 gold bridges there is no trace of artificial cutting or welding. If you are still unsure, then the boxed three-year warranty AMD products are also a good choice. The second is the choice of fan, AMD processor overclocking after the heat problem (Note: overclocking after the heat and not overclocking when the heat is different), has been the most concerned about DIY people, so the choice of a good fan is also crucial.

Intel, on the other hand, has always been known for its stability and better command support for multimedia, making it more suitable for some multimedia enthusiasts, office installations, and home installations for those who are not very computer savvy. From the overclocking point of view, because all Intel processors are locked multiplier, so in the overclocking not show much advantage, although the lock multiplier, but also still can be super, just overclocking the range of smaller, I do not change the case of voltage, will be a piece of P4 2.4 BG super 3.0G, and in some 3D games such as FIFA 2004 can be a stable run, so Intel's! The stability of Intel is still worthy of our trust. Price, Intel's processor can be considered high compared to AMD, although the IT industry pennies a penny, but there is no lack of monopoly factors in it, but it is excellent and stable performance, so that many computer enthusiasts in the installation of the machine, will still be set as the first choice. It is also because of its stability, so many brand-name computers are mostly used Intel's processor, visible Intel's stability is extraordinary. In this way, the Intel processor has an irreplaceable position in a less computer savvy home and business installations. Enough to buy Intel's processor, due to the lock of the multiplier, either bulk or boxed can be assured that the purchase, there will not be like AMD polished products. However, special attention should be paid to the purchase of boxed products, some * merchants tend to use bulk processors with fake Intel fan, repackaged when the boxed product sales, identify the method from the appearance alone is difficult to distinguish, mainly to see the hard plastic packaging inside whether there are traces of dismantling, and then look at the manual is printed or photocopied, counterfeit are generally photocopies. There is also a boxed product can look at the inside of the gift of a small logo (that is, brand-name machines outside the sticker printed Intel Inside of the small patch), genuine small logo thick and hard, there is a layer of hard plastic outside, the fake is relatively thin, with a finger can be scraped off the top of the pattern, and some fakes do not even have a small logo. Now the price difference between a bulk Intel processor and a boxed one is less than a few tens of dollars, and the boxed product also comes with an original fan, so you don't have to buy the fan separately, so buying a boxed product is a good choice.

Second, I personally think that, compared with your CPU and memory and graphics card, your motherboard is obviously a little bit after the strength, you with such a good memory and CPU, it is recommended that you use a higher-grade motherboard, but the NF570 is more expensive, as well as with the NF550 it motherboard choices are a lot of, nvidia aspects of the nforce550 and 570 series. AMD aspects of the 690 series of the recently released. The NF570 is more expensive than the NF570, so why not use the NF550? The NF550 is the most powerful and most powerful motherboard in the world, and the NF550 is the most powerful and most powerful motherboard in the world.

Recommended 550 motherboards:

1 JetBlade HA01GT

2 MSI K9N Neo-F

3 Fantasia TForce 550 SE

4 ARBOR Super Granite AF550T Pro

5 Upscale KN9S

Graphics, 7900GS and 1950GT are very close to each other. 1950GT are pretty close, there's not much difference which one you pick.

The 667 you're talking about would be DDR2, and some of them do go up to 800, but only if they're good quality, like Kingston, ADATA, and Black Diamond.

Three, as for which motherboard it is most compatible with, see AMD's official release

/digi/08/0308/10/46GMLTT1001618J1.html