▷ What is ssd, how does it work and what is it for?

Hard drives, or better said, solid state storage drives or SSDs are here to stay. Almost all users who buy new equipment can find a storage unit of this type inside. But what is really an SSD and how does it work ? In this article we will talk in detail about this electronic element and what it is that differentiates it from the well-known HDD hard drives.

Index of contents

From a few years onwards we have been fortunate to experience great changes in our computers. First were the multi-core processors and their architecture. This led to the equipment becoming faster and faster, improved graphics cards, RAM. But there was still a huge bottleneck in our team and it was none other than the hard drive. With the entire computer full of integrated circuits, we still had a mechanical element inside.

So it would be useless to have an extremely fast processor if the access to the data content was really slow. For this and other reasons, the digital storage industry got to work and consequently lower the costs of creating this new type of units. As its cost decreased, so did the capacity to store data, and also its reliability.

These days, we already have this element practically standardized and common in all new equipment. And at a relatively affordable price. If you want a fast computer, then you must have one of these for your operating system. So let's see what these SSD drives are all about.

What is an SSD

The solid state drive or SSD (solid state drive) is a storage device for data that is based on the use of non-volatile memory or commonly called flash memory. Thus replacing the magnetic disks of traditional hard drives.

These flash memories, successors of the old EEPROM, allow the read and write operations of multiple memory locations in the same operation, thus increasing the speed compared to EEPROM memories, which could only read one memory cell in each operation.

The use of flash memory involves the use of chips to store memory. By eliminating the moving parts of a normal hard disk, we will considerably increase its access and write speed.

In 2010, these reports took another leap, which was the one that really led to the lowering of manufacturing costs and therefore the accessibility of these by users. And it is the use of NAND doors to manufacture these flash memories.

One of the most striking features of a NAND logic gate (AND or inverted Y), is that it can retain data inside even when the power has been cut.

These NAND gates are made using floating gate transistors, which is an item where bits are stored. In the case of RAM memories, these transistors need a continuous power supply to maintain their state, and not in flash memories. When a floating gate transistor is loaded it has a 0 inside, and when it is unloaded it has a 1.

These memories are organized in matrix form, which in turn is formed by a series of consecutive NAND gates. We call the complete matrix block and the rows that make up the matrix are called pages. Each of these rows has a storage capacity of between 2 KB and 16 KB. If each block has 256 pages, we will have a size between 256 KB and 4 MB.

Difference between SSD and RAM

With this, the first thing that comes to mind is RAM memory. As we know, this type of memory is used to supply the data and programs to the processor. When we turn off a computer the RAM memory is completely empty unlike SSD drives.

The difference lies in the use of NAND gates. These logical elements store the last electrical state inside, and also remain even without power supply.

Manufacturing Technologies

Basically there have been two storage technologies for the manufacture of these devices. It started out experimenting with RAM based drives. This required an element that constantly supplied them with energy so as not to lose the data.

Due to these limitations DRAM technology in these units was scrapped with the appearance of NAND gates with non-volatile storage. This is the one currently used and there are three different manufacturing technologies:

SLC or individual level cell

Using this method it is possible to store one bit of data for each memory cell. Its construction is made of individual silicon wafers with which you obtain a thin memory chip and a single level of storage. These chips have the advantages of a higher data access speed, greater longevity and less energy consumption. On the other hand, they have a lower memory capacity, so it will be necessary to build a greater number of sows, therefore increasing their construction cost.

Its manufacturing for now is limited to industrial and server cluster environments where the quality of storage must be superior.

MLC or multi-level cell

This manufacturing method is just the opposite of the previous one. Each memory chip is manufactured by stacking silicon wafers to form a single multi-level chip. As for its advantages are those of greater storage capacity per chip, it is possible to store two bits for each cell, which makes a total of 4 different states. And also a cheaper manufacturing cost.

As disadvantages we quote just the opposite than in the previous case: slower access and chips with less durability.

TLC or triple level cell

In this case, the manufacturing process manages to implement 3 bits for each cell, allowing up to 8 states to be stored. The manufacturing price is cheaper and access to content less efficient. They are, therefore, the cheapest units to acquire, but with a cell life limited to around 1000 writes.

TRIM technology

A subject pending in SSD storage units is precisely their durability. Memory cells degrade for every write and erase performed on them, this causes heavily used drives to degrade quickly leading to file integrity failures and loss of files.

The process of deleting files from an SSD is quite complex. We can write content at the row level, but we can only delete at the block level. This implies that if there are useful files in this block in addition to those that must be deleted, this will also be deleted.

To prevent valid files from being deleted, these files should be taken and saved in a new row, then delete the block and then rewrite the valid data where it was previously. The consequence of all this process is a further degradation of the memory cells by having to do extra writes and deletions.

In response to this, technologies such as TRIM emerge. TRIM allows communication between the operating system and the storage unit so that it is the system itself that tells the SSD the data that it has to erase. When we erase data in Windows, the data is not physically erased, but instead acquires ownership of not being used. This allows to diminish the processes of writing and physical erasing of the memory cells. From Microsoft, this technology has been implemented since Windows 7.

Physical components of an SSD drive

Regarding the components of an SSD drive, we can mention three critical elements:

Controller: is the processor in charge of administering and managing the operations carried out on the NAND memory modules.

Cache: Also in this type of units there is a DRAM memory device to speed up the process of data transmission from the unit to RAM and the processor.

Capacitor: Capacitors have the function of maintaining data integrity when there are sudden power outages. If there is data in motion due to a cut, thanks to the capacitors it will be possible to store this data to avoid loss.

Connection Technologies

SATA

Common SSDs have the same connection technology as normal hard drives, that is, they use a SATA 3 port to connect these to the motherboard. In this way we will have a transfer of 600 MB / s.

PCI-Express

But there is another even faster connection and communication technology called NVMe. Using this method, the units will be directly connected to the PCI-Express expansion slots on our motherboard. In this way it is possible to achieve transfer speeds of up to 2 GB / s in reading and 1.5 GB / s in writing.

As is normal, these hard drives do not have the typical 2.5-inch rectangular encapsulation format, but look like expansion cards like capturers or heatsinkless graphics cards.

M.2

This is the new communication standard intended to replace the SATA type in the medium and short term. It uses both SATA and NVMe communication protocols. These units are directly connected to a specific port located on the motherboard. In this way we avoid occupying PCI-E slots and we will have specific ports. This standard does not have the speed of PCI-E but it is much higher than SATA and there are already drives from all manufacturers at moderate prices.

Aspects to consider of an SSD

When buying an SSD we must know both its advantages and disadvantages, and if our system is appropriate.

File systems

As we have seen the management of an SSD drive is quite different from what we saw for normal hard drives. This is why traditional file systems had the need to update their internal operating structure to suit the needs of these drives. If it were not, it would cause a rapid degradation of the units drastically shortening their life.

NTFS

A clear example is the Windows file system. One of the first optimizations that were implemented, since Windows Vista was to correctly align the partition to the system. This allowed having to carry out extra read and write operations, because the organization of sectors is different in mechanical units and SSD.

In later versions from Windows 7, the systems implement improvements for the SSDs such as the deactivation of the file defragmenter, the Superfetch service, ReadyBoost and the introduction of the TRIM command to extend the life of the SSD.

Advantages of an SSD over a mechanical disk

• Read / Write: Significant increase in basic operations by eliminating mechanical components. Its most appreciable and significant characteristic. Opening of applications and files: directly from the above it follows that the applications and files open much faster and the startup time of the computer will drop dramatically. Failures and security: The time between failures increases considerably and the security of transactions increases by improving data cleaning and there is no variation in performance when the unit is full or empty. Deleting files is also safer, since once physically deleted, files are completely unrecoverable. Energy: it will require less energy consumption and heat production. Noise: since there are no mechanical elements, noise production will be zero. Weight and resistance: by reducing the mechanical components and size, their weight is less and the resistance to shocks is much better.

Disadvantages of an SSD

• Useful life: these units generally have less useful life than traditional discs. This is relative to the intensity of use that is given to these and to the manufacturing technology. Price: The cost per GB is considerably higher than traditional disks. So we find much smaller capacity discs at higher prices. Storage Capacity: Hard drives still exist on the market with less storage capacity than mechanical drives. Not due to hardware limitations (since some time ago it was published that Nimbus Data planned to build a 100 TB SSD) but because of the cost of these. Data recovery: As discussed in the advantages, SSDs delete files permanently, and this is also a disadvantage if what we want is to recover deleted files. Despite this, TRIM technology gives us an opportunity in this regard. Catastrophic failures: while the mechanical disks are gradually degrading and we can notice it, the SSDs fail without warning and this error is total and final. So we will lose the files almost completely safely. Maintenance tasks: In this case, these tasks are very damaging to a hard drive. Defragmenting an SSD does not make sense, but it is not recommended to configure part of the space to virtual memory. This process causes it to wear even more.

If you want to know all the details about mechanical hard drives we recommend our article:

With this we finish our explanation about what an SSD is and how it works. Has it been useful to you? Do you have any questions? ?