ssd 256gb From DDR to DDR4, the memory core frequency has basically not improved much!

ssd 256gb From DDR to DDR4, the memory core frequency has basically not improved much!

harddriveSSD OEMSSD 256GB has evolved from DDR memory in 2001 to today in 2019 ssd 256gb, has gone through DDR, DDR2, DDR3, DDR4.” Four big specs are in the era (DDR5 is also out now). The operating frequency of memory has also evolved from 266MHz in the DDR era to 3200MHz today. This frequency is called Speed in the operating system, equivalent frequency in memory terminology, or simply abbreviated as frequency. The higher this frequency ssd 256gb, the greater the throughput of memory IO per second. But in fact, memory has a most basic frequency called core frequency, which is an oscillation frequency when the actual memory circuit works. It is the basis of memory work and greatly affects the IO latency of memory. I want to reveal another side to you today, this called core frequency has basically not made much progress in the past 18 years.

You can see the speed of the memory on your machine on Linux.

The above command can see the physical device of the memory on each socket, because the result is too long I only extracted one of the memory information column. For us developers, there are two key pieces of data.

All the memory modules on my machine have a speed of 1067 (don’t laugh, because my test machine is a machine that has been eliminated from online warranty, so it is a little older). Multiply the Data Width and Speed to get the data bandwidth. We have summarized the speed and bandwidth of memory at various stages of history, as shown in the figure below.

with Linux we only see one speed in memory, which is the frequency of data transfer. This frequency is also called Data Speed, or equivalent. Various merchants also mark this frequency in a particularly prominent position on the memory sales page, reminding consumers how fast and fast their memory is. But in fact, from the technical parameters of the memory module, there is one of the most important frequencies, which is the core frequency, which is the oscillation frequency of the memory circuit and the cornerstone of all the work of the memory.

let’s take a look at a more comprehensive and detailed picture of each generation of memory.

I have summarized the frequency table comparison of DDR4 memory from the SDR era to the present. You can see that the core frequency has not made substantial progress for many years, which is limited by the limit of physical materials, and the core frequency of memory has been hovering between 133MHz~200MHz. The memory Speed we see is amplified by various technical means based on this core frequency. The reason why we feel that the existence is getting faster and faster is that the amplification technology is constantly improving.

I tried to find commands to check the core and IO frequencies under Linux, but I couldn’t find them, and the various memory sticks on sale seem to rarely mention it. But we are IT practitioners, not ordinary users. Therefore, I think it is necessary for everyone to understand this principle. (In fact, these two frequencies will affect the latency parameters of the memory discussed later, and the delay parameters determine

the real performance of the memory)

> to summarize a sentence, the real working frequency of memory is the core frequency, and the clock frequency and data frequency are amplified by technical means on the basis of the core frequency. The newer the memory, the more magnification there is. But in fact, these amplification methods have some limitations. For example, your memory data storage is not continuous, and DDR2 and DDR3 data prefetching will not help you much. For example, if your process data is stored in a Bank Group, your process memory IO will not reach the speed advertised by DDR4 manufacturers.

In addition to frequency, the memory has several important parameters, but also no command to view is found in Linux. It is not particularly easy to find these parameters on the memory sales page.

All memory modules have four parameters: CL-tRCD-tRP-tRAS. The most important of these are the three parameters CL-tRCD-tRP, and with a little effort, you can find these 3 values in all the memory on sale. For example, the CL values of the classic DDR3-1066, DDR3-1333 and DDR3-1600 are 7-7-7, 8-8-8 and 9-9-9, respectively. Now’s more popular desktop memory Kingston DDR4 2400 8G, its timing is 17-17-17.

the fourth parameter is sometimes omitted. There are two reasons, first: today’s developers don’t need to deal directly with memory, and operating systems are more memory-friendly, and rarely this overhead actually occurs. Second, the value of this overhead is much larger than the other values, which is really not very pretty. Merchants simply avoid talking about it in order to sell more memory.

> OK, here’s the problem. Why does the more memory progresses, the larger the latency period?
This is because the number of labeled delay periods is calculated by dividing the delay time by the memory clock period (Speed/2). This is actually not too scientific, and the most direct way should be to use delay time to evaluate. Latency is largely limited by the core frequency of memory. And the core frequency has basically not improved over the years, so the delay time will not be substantially reduced. Assume that the delay time is constant and the clock period is doubled, so that the delay period appears to be the new memory is larger.

I’ll take you here today, and I’ll use experiments later to let you understand the actual latency of your memory.

Welcome to pay attention to the personal public account “Develop Internal Strength Cultivation” to open up the two veins of theory and practice! harddriveSSD OEMssd 256gb