Timing, RAM and PC performance

Computer terminology is sometimes striking in its complexity. Because of this, the user and, at the same time, the end buyer faces certain selection problems during the purchase of the computer or updating its configuration. One of the important characteristics of the PC is the so-called timing. RAM is characterized as a parameter of the frequency on which it operates, and the size of the latency of access to other modules of the computer.

Before proceeding to the answer to the question, what is timing, we will describe the basic principle of the operation of RAM - random access memory.

How does "operativka" work?

RAM (RAM) is one of the most important parts of any computer. It temporarily stores the data necessary for the operation of the processor. The transfer of information in this case is carried out directly from the memory block to the core or through a special ultra-fast memory. In simple words, RAM is a few microchips that store the data of all programs launched by the user. But is not it possible to store all this on a hard drive, because this is also a memory? Unfortunately no. It's all about speed and reliability. A hard disk is a mechanical device with a low operating speed (in comparison with the needs of the processor) and a limited resource. RAM is devoid of these shortcomings, it is fast, and its resource does not depend on the number of hits.

Classification

There are two types of memory:

• SRAM - static type of RAM;
• DRAM - dynamic type of RAM.

Without deepening, in particular, the technical implementation of SRAM-memory, we can say that such bars are very fast. Delays and data transfer in the RAM block occurs instantly. But, unfortunately, this implementation is expensive. In addition, the volumes of the memory module are limited by the relatively large size of the transistors. SRAM modules are used as ultra-fast cache memory, which is used on processors, hard drives and other PC modules.

Dynamic type of RAM - these are the usual rectangular bars that are located on the motherboard. This memory differs comparatively cheap and large volumes. But its blocks have their disadvantages:

• Since the bar contains capacitors, it is necessary to regularly "regenerate" the charge in them to ensure that the data is not lost. This task is performed by the central processor. But during this access to memory, all operations with it are suspended.
• The speed of such a bar is much smaller than that of a static one.
• An important role is played by properly selected timings. RAM with large volumes and high frequency will not always be able to show the necessary productivity due to high delays.

Types of RAM

At the moment there are only 4 types of memory modules:

• DDR is an obsolete RAM type that is used on very old computers.
• DDR2 - blocks of such RAM can still be found in old PCs in government agencies and educational institutions. The speed of this memory does not allow you to cope with highly loaded modern applications, but it is sufficient for working with text editors and for surfing the Internet.
• DDR3 is the most common memory module. Energy consumption is 40% less than the previous type, and the speed of this memory is much higher.
• DDR4 - the evolutionary development of memory. Such modules are able to fully satisfy all the demands of a modern user. With the optimal configuration, the unit can provide a throughput of 34.1 GB / s.

Memory timings

Now we know what RAM is. Well, what is timing? This is the delay between sending and executing a memory bus command, which is measured in cycles.

DRAM consists of cells, united in two-dimensional arrays. The structure is similar to a lattice, in the nodes of which there are cells. To access the nodes, the controller must know their address, consisting of the line number and the column (coordinates). Individual arrays with the same size of cells are combined into so-called banks.

Thus, the controller first accesses the bank with the row address via the RAS signal. Then, the required row is searched for - it is a cycle of timing RAS to CAS Delay. After that, the controller sends a column number using a CAS signal. Waiting for a response to such a request is called CAS Latency. A timer called RAS Precharge indicates the time between the closing and re-activating commands, Active to Precharge Delay, between the activation and closing commands. Command Rate is the minimum interval between any two commands.

Buying a new RAM bar can very easily determine the timing. The memory is marked with the standard scheme: DDR3 (frequency) CAS Latency - RAS to CAS DELAY - RAS Precharge - Cycle Time, which in reality looks like DDR3 2133 9-12-12-28.

Which is better - fast memory or less latency?

First of all, you need to pay attention to timing. RAM with high frequency can work slowly, because access to the processor is much slower, and therefore this advantage will not be realized. At the same time, delays remain always at the standard level, of course, if you do not manually set the memory timings.

For example, DDR2 1600 6-7-6-18 memory is much faster than DDR3 1866 9-9-9-24. As you can see, in the second case we have a more advanced generation of RAM with a higher frequency, but too many latencies simply neutralize this fact. When acquiring new memory, try to choose a model that has the minimum possible delays. By doing this, you will ensure the high performance of the computer as a whole.