The guide to understanding the RAM of your computer: never run out of memory

Traditionally there has been some confusion when talking about the memory of laptops, desktops and even mobile devices such as smartphones or tablets. “What memory does your computer have?” Is a question that we do not always know how to answer clearly.

Sometimes we confuse gigas with megabytes, the storage memory with RAM , and we get a little messed up. If you are an advanced user, apologize if what we are going to tell now is already known, but in order to undo this confusion we will start with simple concepts.

Computers run programs

The reason for being a computer is to run programs. As much as they look smart, they are nothing without the software . Everything a mobile, tablet, laptop or desktop, Windows or Mac or Linux, with processors of Intel, Qualcomm, AMD, Mediatek or the manufacturer that is, is just that: process program instructions. One after another, from start to finish.

A processor (a CPU or central processing unit), as soon as it receives power when you press the power button, waits for instructions to start executing them . The boot sequences of computers are responsible for initializing the system to run the operating system, on which, in turn, the programs we use in our day to day are executed.

Whether games, Photoshop, Google Chrome or Star Wars Battlefront, it all comes down to a processor that “reads” instructions and processes them according to the structure of the program that is running.




A CPU processes instructions at a clock frequency of several Gigahertz (GHz). The clock of a computer is like the captain of a galley. Mark the rhythm to which instructions are processed, in the same way as the captain marks the rhythm to the rowers . The nuance is that, if they had to paddle according to the clock frequency of a computer, they would have to move the oars a billion times per second, assuming that the clock frequency was 1 GHz.

That is the rate at which current processors work. That is, assuming a frequency of 1 GHz, one instruction is processed every nanosecond . At 3 GHz, 3 instructions would be processed for every nanosecond, and so on.

So that there are no bottlenecks that slow down the operation of the CPU, the processors have to receive instructions at that rate . Otherwise, they would have to be waiting for them, which would not make sense. Now, where are the programs that contain these instructions stored? The answer is: in memory.

Simple, but in practice, things get complicated. For example, there are several types of memory. And the instructions and data are moving from one to another as appropriate to make the processor does not have to wait to execute the code sequences. The faster this memory is, the more expensive it will be and the less quantity we can have .

So the engineers had no choice but to design a hierarchy of “containers” (buffers or cache memory) in which small amounts of super fast memory are used in a first level, larger amounts of memory somewhat less fast in a second level, even greater amounts of cache even less fast in a third level, to reach a fourth level of RAM , which is the most familiar to users. The memory of the first three levels is called “cache” and is integrated within the silicon of the CPU. In a fifth level would be hard drives, SSD or magnetic.

To give you a rough idea, the RAM is what we can buy in stores , while the cache memories are part of the electronics of the CPU. Another difference between the cache and RAM is its technology: cache memories are of type SRAM (Static RAM), very fast but very expensive, while what we know as RAM is DRAM (Dynamic RAM), less fast , but less expensive.

If we talk about sizes, the L1 cache moves in values ​​of the order of 32 KB, memory L2, of the order of 256 KB and memory L3, in the order of 3 – 12 MB. That is, very small quantities compared to the ones we usually use for RAM, in the order of GB . One MB is 1,024 KB and 1 GB is 1,024 MB, so you can get an idea about how little cache memory is available on computers.

As we have already mentioned, SRAM memory is faster, but very expensive. RAM is cheaper, but slower. Faster, lower capacity, and more price .

Volatile, non-volatile memory and hard disks

Cache and RAM memories, on the other hand, belong to a wider category, that of volatile type memories . That is, if we turn off the computer, its content is lost. For this reason, it is necessary to have non-volatile memory to store the data and programs.

Hard disks and SSDs are non-volatile memory, and this is where a computer comes to boot to load in memory the instructions and data that will run during its operation . Non-volatile memory falls into the category of storage systems.

So, when they ask us about the memory of the device or the phone or tablet, they will usually refer to the volatile RAM memory . If you ask about the storage capacity, we are talking about non-volatile memory. The hard drives. The problem with non-volatile storage systems is their slowness. Access to content on a hard disk or even on an SSD is very slow compared to accessing RAM .

Doing some numbers, if we take as reference the time that a clock cycle lasts, that for a processor at 3 GHz is about 0.3 ns and if we normalize the scale to make the comparison reference is 1 second , we would have the following:

 

  • accessing the L1 cache is about 3 seconds
  • access the L2, about 9 seconds
  • access the L3 cache, 43 seconds
  • when we move to RAM, we are talking about a 6-minute access time
  • for an SSD disk, between 2 and 6 days
  • access to a magnetic hard drive, is between 1 and 12 months

 

In denormalized “computer” time, those 12 months are about 10 ms (milliseconds). A manageable time, but it is a figure much greater than the 0.9 ns (nanoseconds) it takes to access the first level L1 cache. And an unacceptable time for normal use of a computer (or a mobile or tablet).

The importance of RAM

The RAM, ultimately, is the memory that combines a decent size with an affordable price . It is the memory where our programs and data are “loaded” from the SSDs and hard disks so that they are “on hand” when they are claimed by the processor prediction algorithms, so that they are loaded into the caches before they The CPU claims them.

In this way, the CPU just has to wait, since the data and instructions are organized so that they are always in the fastest memories. If we do not have enough RAM, or the prediction algorithms fail, this flow is broken and the processor will encounter the harsh reality that is facing the slowness of the hard drive or the SSD .

For example, if the processor needs a data that is not in the caches, or in the L1, L2 or L3, it will go to RAM. If it is not in the RAM either, it will have to go to the hard disk or the SSD to read it , and therefore we will have to wait enormous amounts of time for the reading operations to take place.

The more RAM we have, the more likely we are that there is no need to look for data or instructions for slow non-volatile storage systems .

When we say that “the computer is slow”, in many cases these are situations in which the system works more time with RAM and hard disks than with caches . In this order of things, it is interesting to say that the best friend of RAM is an SSD , especially when we turn on or off the computer, or when we start a new application on the computer.

At times like turning on, turning off, or when launching a new application or reading a photo or text file, the system has to work “yes or yes” in the area of ​​RAM and non-volatile storage systems . And, as we have seen, using an SSD disk instead of a magnetic one, is the difference (on a normalized scale) between taking days to perform data access operations and programs, taking months.

When using SSDs, starting and stopping the computer or, in general, loading applications into RAM, are tasks that are performed in a fraction of the time it takes with magnetic disks . And if we suspend the equipment, the boot is even faster, because in this case you do not have to empty the RAM to save it on the disk, but it maintains its content thanks to the power of the volatile memory is not cut off.

Of course, not to use, say, 32 GB of RAM, the system will go faster. There is an amount of memory below which we will have problems, but above certain limits, unless we use very demanding applications, we will not notice difference between, say, 16 GB and 32 GB.

Types of RAM

Memory is usually categorized according to its technological generation. The memory used in desktops and laptops is DDR (Double Data Rate) and the most recent is the DDR4 type, although the easiest you will find in the equipment and stores is the DDR3 type . There is a variant of DDR3 of low consumption, the DDR3L, that you can also find for sale. The different generations are incompatible with each other, so in a DDR3 compatible device you will not be able to install DDR4 or DDR2, for example.

The memory is marketed in the form of DIMM or SODIMM modules . The first are for desktop computers, and the second, shorter, for laptops. The modules have capacities that vary between 2 GB, 4 GB or 8 GB as usual, although there are also 16 GB or even 32 GB. There is also (in general) the possibility of installing a variable number of them, depending on the number of available slots, which can range from 2 to 8 in the equipment with which we are accustomed to work in the domestic field.

Anyway, the trend (for the moment in laptops) is to make RAM integrated in the board, without possibilities of expansion . In fact, another categorization of the types of memory that begins to be relevant, from the perspective of the buyer, is that which refers to whether the RAM is soldered on the computer’s board, or installed using removable modules.

In addition to the type of module and the capacity, the RAM has another variable, the frequency in MHz to which it can work . There are standards that define a set of speeds at which they have to operate under the appropriate standard, but manufacturers usually manufacture modules capable of operating at higher speeds.

Actually, more MHz does not mean that the performance is higher. At higher clock speeds, other parameters that negatively affect performance, such as access latency, are also higher, so in the end, it is usually working with similar effective speeds for the transfer of data between the processor and RAM.

The advantage of having modules capable of running at a faster speed is that they will not “hang up” if we tune the system to work faster . Overclocking is to increase the speed of operation of the components of the system, and if, for example, we increase the frequency of the processor, it will also increase the frequency in the memory buses. If we had a module that worked only at the nominal speed, it would “hang” under those conditions.

What is that about the channels?

In desktop computers, where you can install four and up to eight memory modules, it is usually indicated that you have to install the memory in a certain order. The reason lies in the way in which the CPU accesses the memory. To improve the transfer speed, the processor is able to use several channels in parallel when accessing the memory to increase the number of bits it reads in each read operation .

To do this, it organizes the slots into channels, so that, in a configuration of four slots, two will be for channel 1 and two for channel 2. And the thing is that the modules are installed so that the processor reads the data in memory, taking advantage of the width of the entire channel.

In practice, the performance of the system will not be particularly affected by the non-use of all channels. The speed of data transfer is reduced, but we are talking about figures in the order of tens of GB / s (Gigabytes per second) , enough for there to be room to spare when transferring data to caches.

In laptops, generally, we only have a single memory module, wasting the multi-channel configuration, and this does not result in a significant performance drop . Laptops from 15 ” up, and some from 13 ”, can use SODIMM memory modules in configurations of two, although it is increasingly common to save space by eliminating these slots.

How much memory do I need?

Today, and taking into account the prices of RAM (between 6 euros and 10 euros per GB, by way of approximation), the “sweet” point would be in the 8 GB for laptops and desktop . It is a sufficient amount to move the operating system and work with daily applications and multimedia content that does not go out of the ordinary. Better with an SSD, of course.

With 8 GB we can work with video and photo editing applications in domestic projects. As well as manage in multitasking scenarios without problem. Below, we start having problems if we work with high-resolution image files, high-resolution and quality videos or with many tasks at once . These are not irresolvable problems, since the system will go to “simulated” memory on the hard disks to work (the paging files), but at the cost of slowing down the response to very undesirable limits.

 

Above 8 GB we enter a field with many possibilities for enthusiastic and demanding users . In professional systems, RAM is installed in larger quantities even than the size of our SSD hard drives. In these scenarios, databases and spreadsheets with gigantic sizes are handled, which have to be in RAM to meet the requests of thousands of users at the same time, for example.

In the domestic environment, from 16 GB we can play with the use of RAM, for example, as a place to store temporary system files or applications, which would otherwise be installed on hard drives or SSDs .

You can also install larger amounts of 32 GB or 64 GB and use part of the RAM to configure it as a hard drive in RAM . If we install a virtual machine in virtual units in RAM, the result is spectacular in terms of speed and response.

The amount of memory is especially relevant if we buy a piece of equipment that has RAM soldered, and that therefore can not be extended. They are usually portable Ultrabook or convertible, ultralight and very thin design. The most economical models usually only equip 4 GB , which allows to work in a decent way with Windows and Mac, but it remains a fair enough for demanding and enthusiastic users in order to amortize the investment or to protect themselves from the programmed obsolescence as the Applications and operating systems are designed for more ambitious configurations.

DDR4, is it better than DDR3?

The newest in RAM is DDR4. With the new Intel sixth-generation processors, compatibility with DDR4 RAM arrives in both desktops and laptops . The standard speed (SPD) of the modules goes from 1,600MHz to 2,133MHz and the memory modules reach higher capacities with less technological complexity, so it is not complicated to find 8GB and even 16GB modules .

Anyway, the speed is not really an improvement, since the internal latencies are greater than in DDR3 and in a certain way it leaves “what is eaten for what is served” . The advantages are also in the voltage part, lower than in DDR3, which saves energy, especially in laptops.

So DDR4 is a good argument, of choice (almost) required, everything is said, if you buy a computer with new Intel sixth generation processors, but do not tell you that it is better to buy a computer with DDR4 instead of DDR3 because the RAM performance is better . Or put another way, that memory alone is not the purchase argument of a new computer.

Those who practice overclocking, they should pay attention to the working frequency of RAM, although it is a type of audience that knows exactly what you need and how to configure it.

Better with SSD

As a tip, to say that installing an SSD is a resource that can greatly improve the operation of a computer . As we have seen, it is inevitable that the system has to “fill” volatile memories with the content of non-volatile (hard drives) during such usual tasks as opening an application, or turning the computer on and off.

SSD disks are many orders of magnitude faster than magnetic disks , so we will reduce the temporary penalty of having to search RAM on data or programs.

The same applies in the case that we have enough RAM to configure a virtual disk. When saving or recovering the content of the same to turn off the computer, it will take much less to do it if we have an SSD disk as a storage unit.




The near future

The organization of memory has hardly changed for decades. The most notable advance was many years ago: that of solid-state memory applied to SSDs . But the non-volatile memory was still very inferior in performance to the volatile, which forces us to establish memory hierarchies like the current one.

But Intel has developed together with Micron, a manufacturer of memory, a new non-volatile memory with performance comparable to that of the volatile RAM . This technology, called 3D XPoint, opens the doors to eliminate the barriers between RAM and SSD disks, unifying the difference between both storage subsystems.

Leave a Reply

Your email address will not be published. Required fields are marked *