Why is AMD s clock speed low, but its gaming performance is better?

The architecture is not the same, and it cannot be compared with the actual frequency.

Generally speaking, the higher the monthly performance of the CPU with the same core.

Not necessarily with different cores.

For example, Intel's Ben 4 and AMD's Athlon64, although Intel's processor frequency far exceeds AMD's, but due to too many core architecture pipelines, the efficiency is very low and the power consumption is very high. And AMD's, although the frequency is not high, has fewer assembly lines and high efficiency, so 1* GHz Athlon-64 performance is sometimes stronger than 3

GHz Ben 4.

The main thing is the integrated memory controller, so with less time to access the memory, the game naturally has a good speed.

However, the clock frequency doesn't say anything, and it's something to correct. After CD overclocking, the main frequency reaches more than 4G, and the AM2 3000+ overclocked to more than is already better than CD in games. The CPU sometimes looks at the instructions, threads, caches, execution levels, etc. to illustrate its efficiency in the use of games or software, and sometimes the importance of the external frequency is more critical than the main frequency.

Therefore, the low frequency does not mean anything.

That's what I look like with that model.

The Centrino and the next Core have lower clocks and better performance.

For example, I took a Centrino laptop and easily won the Athlon 3000 or P43G.

The architecture is more advanced and the performance is high, unlike Intel's high-frequency and inefficient, so the AMD processor with a nominal PR value has strong performance.

AMD and Intel are two CPUs, and their main frequency algorithms are different.

The main thing to look at is the cache, and here is the bottleneck.

The frequency and number of cores are not the same as performance, but also depend on the core architecture and design purpose

In a nutshell, Intel core single-core performance is high, and AMD multi-core is cost-effective.

It is more cost-effective to equip computers with AMD processors (including APUs) for home entertainment, and it is recommended to pursue high performance with FX8 cores and i5 or above.

Home processors are roughly ranked in terms of performance

RTR R7 8th Gen i7 R5 8th Gen i5 i7 E3 i5 FX8 i3 FX6 Phenom 6 Athlon4 A8 A10 Pentium G A6 A4 ......The worse ones have been eliminated (i9, i7ee, etc., faith price detection u is automatically ignored).

Xeon is designed to be used as a server, and Lao Zhiqiang is mostly a low-frequency multi-core, so the game performance is not good. The main frequency of the new Xeon (such as after E3-1230) is not low, and the operation effect is stronger than that of i5 and catches up with i7 in terms of supporting more and optimized programs; However, if the software does not support multi-core optimization, then the effect of E3 is comparable to or even slightly inferior to i5. The same is true for FX8 cores, which also require software support for multi-core optimization in order to give full play to actual performance.

A lot of old software doesn't use more than 4 cores, so the FX8 core is as powerful as the E3.

In the final analysis, the actual performance is fundamental, and the main frequency and number of cores are only parameters for comparison within the same system, and they do not represent performance and are not comparable after cross-series. And the processor with the same level of performance, as long as the software optimization is no problem, the operation effect is also comparable. As for multi-core optimization, this is already a necessity of software development, and with the rollout of DX12, the comprehensive performance of multi-core processors will be even more brilliant.

To put it simply, if you take a car as an analogy, if the motor is junky, no matter how high the speed is, it can't outpace the low speed of a high-end motor, right?

Rotational speed = CPU frequency Motor level of advanced = CPU process architecture process and other parameters.

AMD's CPU process and architecture are too far behind Intel for current games, especially online games, and the optimization is too poor, so an AMD flagship FX9590 that is overclocked to 5G is still playing games by Intel Xeon E3 1231V3 (main frequency) that cannot be overclocked.

And the current games rarely have 8 cores and above optimization, and generally only use dual-core 4-core 4-8 cores, so AMD's high-end 8-core CPU cores are basically idle.

Another point,CPU main frequency, just looking at the main frequency, in fact, basically meets the demand,Only a small number of games and online games are sensitive to high main frequency Most large stand-alone overclocking can not increase the number of frames,In the case that the original main frequency has reached the standard。

The CPU for playing games should have a high frequency and a good cache capacity.

To play games and choose a CPU, you can view it from four aspects: the main frequency of the CPU, the bus, the L1 cache, and the L2 cache

1. The main frequency, that is, the clock frequency of the CPU, simply put, is the working frequency of the CPU. Generally speaking, the number of instructions completed in a clock cycle is fixed, so the higher the frequency, the faster the CPU. However, due to the different internal structures of various CPUs, the performance of CPUs cannot be completely summarized by the main frequency.

As for the external frequency, it is the operating frequency of the system bus; The frequency doubling refers to the multiple of the difference between the CPU external frequency and the main frequency. It is expressed by the formula: main frequency = external frequency and frequency doubling.

2. The memory bus speed or system bus speed is generally equivalent to the external frequency of the CPU. The speed of the memory bus is very important for the performance of the whole system, because the development of memory speed lags behind the development speed of the CPU, in order to alleviate the bottleneck caused by memory, there is a L2 cache to coordinate the difference between the two, and the memory bus speed refers to the working frequency between the CPU and the L2 cache and memory.

3. L1 cache, which is what we often call L1 cache. The built-in cache in the CPU can improve the efficiency of the CPU. The capacity and structure of the built-in L1 cache have a great impact on the performance of the CPU, but the cache memory is composed of static RAM, and the structure is complex, and the capacity of the L1 cache cannot be too large when the CPU die area is not too large.

Caches with a writeback structure. It provides caching for both read and write operations. Caches with a write-through structure are only valid for read operations.

Write-back caching is basically used in computers above 486. Among the currently popular processors, the Pentium and Celeron processors have 32KB of L1 cache, the Pentium 4 is 8KB, and AMD's Duron and Athlon processors have up to 128KB of L1 cache. But the performance of the CPU is not judged by the main frequency, it can only be said that the higher the main frequency, the better if other parameters are the same.

4. L2 cache, refers to the cache of the second layer of the CPU, the first to use the L2 cache is the Pentium Pro processor, its L2 cache and CPU run at the same frequency, but the cost is expensive, the market life is very short, so the L2 cache of the Pentium II runs at half the CPU frequency. The next Celeron processor uses the L2 cache running at the same speed as the CPU, and the now popular CPUs, whether it is AthlonXP and Pentium 4, have L2 caches running at the same speed as the CPU. In addition to speed, L2 cache capacity will also affect CPU performance, the principle is that bigger is better, now the mainstream CPU capacity is the largest is 3MB, while the L2 cache with CPU on servers and workstations is higher, several times higher.

Clock speed is not the only factor that determines CPU performance.

And you are also mistaken, now the core era AMD processors are clocked at a higher frequency than Intel. In the Ben 4 era, the main frequency of AMD processors is lower than that of Intel.

The processing power of the CPU at the main frequency and the number of instructions processed per clock cycle efficiency.

The main frequency will directly affect the processing power of the CPU, but just like in the formula, there are two other factors that will also affect the processing power of the CPU, so it is not necessarily that the CPU processing power with a high frequency is necessarily strong.

The P4 has a very high clock speed, but its performance is not as good as that of the Core, which is much lower than it, because it executes too few instructions per clock cycle (lower than the Ben 2) and its long pipeline makes it inefficient (it will process a lot of invalid instructions).

In addition, in addition to computing power, the CPU has other aspects that will affect performance, such as data access efficiency (related to the size of the cache and the channel bandwidth of accessing memory).

The number of instructions per clock cycle and the efficiency of AMD processors are much higher than those of P4 and weaker than those of Core.

Therefore, AMD's processor performance can reach the level of Pentium 4, and the main frequency can be much lower than it, and the main frequency must be higher to reach the core level.

AMD has always been good at architecture, so it doesn't rely on frequency. It wasn't until the latest Cool came out that it was a little overwhelmed.

In addition, AMD's connection architecture is also better than Intel's, so AMD never depends on cache size, and Intel has to cache large capacity for the CPU to make up for the shortcomings of the connection architecture.

In addition, some brothers answer the question of manufacturing process and process, which is not AMD's competitive advantage at all, on the contrary, it is AMD's biggest disadvantage, AMD has been lagging behind Intel in terms of manufacturing process and process (now AMD's latest Phenom CPU only has 65nm manufacturing process, while Intel's Penyn has reached 45nm).

Under the same architecture, the higher the frequency, the better. It's not that the same architecture is not comparable. Like Intel's earliest Core 2 Duo E6300 only has the main frequency, but the performance is much more than the Ben 4, which means the same thing.

The quality and stability are almost the same,AMD was said to be unstable because of the big problem of heating,Quality-speaking, I've used AMD so much CPU has never been broken,There is man-made damage,In the past, AMD's performance was not as strong as Intel,Now AMD has come out of the Ryzen generation and the second generation can basically tie with Intel,So there is no difference in quality and stability,It can only be said that the single core of Intel CPU is really strong,

AMD's CPU is actually not bad, Intel is not strong, Intel is better than AMD in support of professional software, and there is no sense of performance for AMD at home, including the gaming experience.

Intel's focus on the instruction set SSE2 and SSE3 and the very important rendering part, the optimization of professional software is significantly higher than AMD, AMD** performance, and home, the gaming experience is not necessarily worse than Intel.

Now in the game Hardware Killer Lonely Island 3, the CPU horizontal evaluation (about the evaluation, I have always thought that you should not look at the domestic evaluation, the moisture is too high), and the comprehensive performance of 8350 is higher than that of 2600.

In addition, 8350 is slightly higher than 3570 performance, 3570 and 3770 are a hyper-thread, the highest performance improvement of hyper-threading definition is 1-30%, I use it myself, I want to tell you that 3770k is -4%-20% performance improvement compared to 3570, in some cases, 3770k is not as good as 3570k, hyper-threading is not used in all cases, and it is amazing that you can play 10% of the performance of hyper-threading in the game. Hyper-threading is for professional software, industry software.

In addition, AMD has not introduced hyper-threading technology, if 8350 is added to hyper-threading, it can catch up with 3770.

When the AMD desktop CPU is not as good as Intel's, this is not controversial, but it is not as exaggerated as the domestic ** and the evaluation said.

For assessments, there is no income for assessments, they are fair and just at the beginning, and then you start to become famous, you slowly start to have income, and you start to lose your objectivity and fairness.

Upstairs doesn't have such a black AMD, right?,For example, I'll give you an example.,When Intel is still desperately trying to increase the frequency of Pentium4,P4's can't compare to Athlon's.,Now the performance of the CPU doesn't depend on the frequency.,The main thing is to see the architecture is very supportive of technology.。

The main thing is to look at the architecture, and besides, there is no comparison between the two at all, AMD's Even if you have an octa-core processor, you can't compare to Intel's quad-core. That's it.

Because AMD is an agricultural enterprise, it mainly makes farm tools, and CPU is just a side business. A farmer, of course, likes high power and high horsepower, but the efficiency is not high.

Yes. Because the comparison usually requires the same performance, and under the technical level of Intel, the performance of a single core in the multi-core processor is very strong, and AMD is obviously insufficient in this regard, so it can only take the method of either increasing the running frequency or increasing the number of cores to improve the performance of the processor, so it leads to the phenomenon of AMD's processor "fever", and the running temperature is too high, which leads to the instability of the processor performance on the one hand, and shortens the life of the hardware on the other hand, and also increases the consumption of electric energy.

Therefore, Intel's CPU is very expensive, and AMD has to reduce the cost performance route. But after all, AMD can persevere, otherwise Intel would have suffered a lawsuit under the monopoly law, so it seems that the existence of a competitor is actually to help him avoid monopoly and improve his identity, why not.