What s the use of CPU cache so small?

Extremely fast and costly.

Level 1 capacity increases and slows down, including memory.

Put the most likely to be accessed in the nearest tell-the-cache.

The update is fast, it's all binary.

It's an instruction, not a program... Cache, memory, hard disk, one is bigger than the other, one is slower than the other, and one is more expensive than the other.

Categories: Computer, Networking, >> Hardware.

Problem description: Does anyone know? Tell me, thank you.

Analysis: When the CPU executes instructions, it will put the execution result in a component called "register", because the "register" is integrated inside the CPU, and the important components of the CPU such as ALU, so the instructions in the register are quickly accessed by the CPU, but after all, the capacity of the register is too small, and a large number of instructions and data required by the CPU are still in the memory (RAM), so the CPU needs to frequently send and receive instructions and data to the memory in order to complete the instruction operation.

Since the processing speed of memory is much lower than that of the CPU, the traditional system bottleneck is created here, and the CPU often spends a lot of time waiting for the memory to prepare for the processing of instructions.

In order to solve this problem, people have integrated a much faster "cache" than memory inside the CPU, which is the earliest "cache".

L1 cache is a memory that runs completely synchronously with the CPU, which is what we often call the first-level cache, if the data and instructions required by the CPU are already in the cache, then the CPU does not have to wait, and can directly obtain data from the first-level cache (L1), if the data is not in L1, the CPU extracts data from the second-level cache (L2), which greatly improves the work efficiency of the system.

What is caching technology:

The English word for cache is cache. A special memory subsystem in which frequently used data is replicated to facilitate quick CPU access. The cache memory of the memory is stored for frequent access.

The contents of the RAM location and the address where these data items are stored. When the processor references an address in memory, the cache checks to see if it exists. If this address is stored, the data is returned to the processor; If the address is not saved, regular memory access is performed.

Because the cache memory is always better than the main RAM

The memory is fast, so when.

When the access speed of RAM is slower than the speed of a microprocessor, cache memory is often used.

What the caching does:

Before the CPU starts executing any instruction, it first retrieves that instruction from memory, along with some other relevant data and information. To speed up the operation of the CPU, almost all chips employ two different types of internal memory, namely cache. Caches are used to temporarily store frequently used pieces of program or data.

L1 cache is the best-performing type of cache, and the processing unit that interprets instructions and performs arithmetic operations forms the core of the CPU. The CPU can read instructions or data stored in the L1 cache at full speed. Intel's processor products typically have 32K L1 cache, while competitors like AMD or Via use more L1 cache.

If the required instruction or data is not found in the L1 cache, the processor looks at the L2 cache with a larger capacity. L2 caches can be either integrated into the CPU chip or used as an external cache. pentium

The II processor has 512K L2 cache and works at half the speed of the CPU. Celeron as well as newer pentium

The iii chip has 128K and 256K on-chip L2 cache, respectively, and is capable of running at full processor speed.

For instructions or data stored in a slower L2 cache, the processor often has to wait 2 to 4 clock cycles. In order to make the most of computing resources, the CPU can view and execute other instructions that are waiting to be processed during this time, but do not require additional data, increasing the speed of the entire system and reducing idle time to a minimum.