The arrival of the first solid-state devices or SSDs a few years ago, represented a revolution in the entire set of internal storage in PCs. A revolution that could only be compared to the jump from the old punch cards to the first hard drives . But surely there are more than one of you wondering how an SSD can work so fast. That is what we are going to explain in this new article.
Externally, an SSD and a 2.5-inch hard drive are very similar in their form factor and connectors. However, this is where all the similarities between the two storage devices end .
The biggest difference between a solid device and a mechanical one is in how the information is stored and accessed. In a hard disk the information is stored all in sectors distributed on the surface of a plate that rotates at high speed (5,400, 7,200 or up to 10,000 rpm). To access this information (or record more of it) a reading head is used, which locates it on the surface. However, an SSD lacks moving parts in its design.
The only thing that is inside a solid state device are chips and of course a PCB that connects and joins them . NAND Flash memory chips. Cache chips. And a memory controller.
It is precisely in this simplicity where the great advantage of SSDs over hard drives lies .
Access to information on SSD memory chips is faster
For a hard disk to locate information inside it, it must turn the plates to where the data has been stored, until it reaches the sector or sectors where they appear. But, as you well know, in very few occasions the information is stored sequentially.
Instead, it is usually fragmented between different sectors of the plate, so that, to access complete data, it is very possible that the hard disk must search between several different sectors. Sectors that can be located in very different places from it and therefore everything is complicated.
All this generates latencies in the searches themselves, not to mention, also, that the process of accelerating the plates of the hard disk so that it can read the data, which also takes time. In fact, the access times for this type of storage devices are measured in milliseconds, which although it may seem an incredibly short time, in an HDD versus an SSD is really too much.
However, solid devices do not have any of these drawbacks. The moment a request for data reaches the internal controller, it can retrieve the data almost instantaneously. Regardless of whether they are distributed in different NAND Flash memory chips.
That is why defragmenting the information on an SSD hardly brings any benefit to the performance of these devices. And, in fact it can be negative for NAND Flash memory cells because it wears them down unnecessarily. In general, Windows 10 by default already defragments even the SSDs automatically and at most once a week, since defragmenting in a massive way only benefits the HDD and harms the SSD.
The fact is that if the latency of a hard drive is measured in milliseconds, that of an SSD is measured in microseconds, hence we speak of near-instant access as such.
In fact, the biggest thing that can slow down the performance of a robust device is the bus itself to which it is connected. The SATA bus is the most common for an SSD since they came on the market. Only the most modern models are now capable of using the motherboard’s PCIe bus, NVMe, with the consequent increase in performance.
