What does the CPU s L2 cache have to do with memory?

This also reduces the opportunity to read page files in the system disk, and reduces the pressure on the system disk, and the maximum value cannot exceed the remaining space value of the current hard disk.

The working frequency is not the same between the CPU and the memory. The CPU is many times higher than the memory, in order to solve the problem of the difference in the rate of exchange of information between the two. Cache appeared.

But now one buffer is no longer enough, so another one is added. This is what L1, L2 caches roughly mean.

Of course, the larger the cache, the faster the information can be processed.

Level 1 cache: L1

cache。It is integrated inside the CPU and is used for the temporary storage of data during the processing of data by the CPU. Since the cache instructions and data work at the same frequency as the CPU, the larger the L1 cache capacity, the more information can be stored, which can reduce the number of data exchanges between the CPU and the memory and improve the computing efficiency of the CPU.

However, because the cache memory is composed of static RAM and the structure is complex, the capacity of L1 cache cannot be too large in the limited CPU chip area.

L2 cache: L2

cache。Due to the limitation of L1 cache capacity, in order to increase the computing speed of the CPU again, a high-speed memory, that is, a L2 cache, is placed outside the CPU. The main frequency is flexible, and can be the same frequency as the CPU, or it can be different.

When the CPU reads data, it looks first in L1, then from L2, then in memory, and then in external memory. Therefore, the impact of L2 on the system cannot be ignored.

L1 can't be too big.

L2 is big, good.

However, the cache consumes a lot of power.

And it also takes up a lot of area.

It's not too big.

Some CPUs only have L2

In some high-end CPUs (like Intel's Itanium), we often hear about L3 Cache, which is designed for data that misses after reading L2 Cache—a kind of cache, in CPUs with L3 Cache, only about 5% of the data needs to be called from memory, which further improves the efficiency of the CPU.

There is an L3 and there must be an L2. Of course, it's good to have L3. But expensive. It depends on the architecture.

You can't just look at this to judge the strength of the CPU.

1.Regarding the main frequency of the CPU, it is the frequency of data processing, that is, how much data can be processed in a cycle. Of course, the higher the better, but different series of CPUs cannot be compared, for example, AMD's 3000+ is used with Intel's P4

Rivaling, and 3000+ only, for example, the current core 2 main frequency is compared with the previous P4, the main frequency is doubled, but the performance is doubled.

2.For example, if the computing speed of the two CPUs is the same, the bus is different (one 533, one 800), the bus is like a road, the wider the road, then the more data can be transported to the CPU in a period of time, that is to say, it is useless that the CPU operation is fast and the bus cannot catch up with the progress.

3.Regarding the L2 cache, the bigger the better, the CPU will store the data that is often used in the L2 cache, and the data in the L2 cache is called faster when the CPU is operating than the data in the memory, so as to improve the overall operation speed, but the space of the L2 cache is first, and when it is full, there is new data to be put in, and the CPU will choose the least commonly used to exclude, so the size of the L2 cache is also as large as possible.

Haha, I'm so tired, I typed so many words, I hope to help the landlord.

ps: Despise pulling articles from the Internet.

CPU clock speed, bus frequency, and L2 cache.

The bigger the computer, the higher the performance.

CPU main frequency, bus frequency, frequency multiplication.

The main frequency is not the higher the better, the bus frequency is one grade higher than the CPU, and the L2 cache has a decisive effect on the speed of running the program.

Big, of course. The larger the L2 cache, the higher the hit rate of the CPU accessing important data in the cache, which will reduce the number of times to retrieve data from low-speed memory, and the natural operation will be faster.

Of course, it's not like bigger is better. Take the early 775 CPUs as an example:

Celeron E1500, Pentium E2180, Core 2 E6300, Core 2 E6320 The cores of these CPUs are almost the same, and the difference is mainly reflected in the L2 cache. 512K for the E1500, 1m for the E2180, 2m for the E6300 and 4m for the E6320. If their main frequency is tuned to the same value, the E2180 performs significantly better than the E1500, and the E6300 significantly outperforms the E2180.

However, the improvement of the E6320 relative to the E6300 is not so obvious.

This shows that the initial increase in L2 is effective, but when it is large enough to increase its capacity, the contribution to the performance improvement is not so large. At this time, the capacity of L2 is no longer a bottleneck, but the core computing power has become a bottleneck. For example, L2 is like a warehouse, when the original warehouse of a factory is very small, it is beneficial to increase the production capacity by transforming and increasing the warehouse, but the warehouse is large to a certain extent, and it is useless to increase it, because the production line has not increased, and the warehouse has been idle.

In addition, too large L2 will increase the complexity of the CPU, increase the number of transistors, increase power consumption, and reduce the yield rate.