The read and write speed of mechanical hard disks is determined by the speed, data density, cache size, interface type and seek time. High speed shortens latency, high density improves linear speed, large cache optimizes random read and write, while fragmentation and background programs slow down the actual experience.

The reading and writing speed of a mechanical hard disk is ultimately determined by several core elements. The most intuitive feeling is the disk speed, data storage density, and cache size. Of course, the hard disk interface and internal seeking mechanism also play a role that cannot be ignored. These factors intertwined to ultimately constitute the performance we actually experience.

Solution

To deeply understand the reading and writing speed of mechanical hard disks, we need to focus on several key technical points. First and most importantly, the disc speed (RPM) . This directly determines how much data the head can read or write in unit time, and how long the seek time is. High speed means data passes faster under the head, thereby improving data transmission rates.

The second is the data density (Areal Density) , which is the amount of data that can be stored per unit area. At the same speed, if the data density on the disc is higher, the head can read more data by passing a short distance, which will undoubtedly increase the linear read and write speed. It's like on a highway, the more lanes the more vehicles you can pass within a unit of time.

Cache (Cache/Buffer) is also an important accelerator. A piece of DRAM is usually integrated inside the hard disk as a cache, which is used to temporarily store data that is about to be read or just written. When the data requested by the system happens to be in the cache, the hard disk does not need to perform time-consuming mechanical search operations and returns directly from the cache. This is very effective for small files, random read and write, or frequently accessed data.

In addition, the interface type also limits the upper limit of data transmission between the hard disk and the motherboard. For example, the theoretical bandwidth of SATA 3.0 is 6Gbps. Although mechanical hard drives rarely run over this bandwidth, an outdated interface will undoubtedly become a bottleneck.

There is also Seek Time , which refers to the time it takes for the magnetic head to move from one track to another. It is limited by the design of the robotic arm, the performance of the drive motor, and the optimization of internal algorithms. The shorter the search time, the better the random read and write performance.

File fragmentation is also a factor that cannot be ignored. If the file is stored in fragmented on the hard disk, the head needs to be moved frequently to read a complete file, which will greatly increase the search time and thus reduce the actual read and write speed.

Why do high-speed hard disks feel smoother in actual use?

I personally think that the main reason why high-speed hard disks can bring a smoother experience is that they directly shorten the "waiting" time. Think about it, a 7200 RPM hard disk has a disk that rotates 120 turns per second, while a 5400 RPM hard disk only rotates 90 turns. This means that no matter where the data is, the high-speed hard disk can "turn" the target data area under the head faster.

This is not just a theoretical increase in data transmission rate. In practical applications, such as if you start a large game or load a complex video editing project, the hard disk needs to read thousands of scattered small files. At this time, the head needs to jump frequently on the disc. The high speed reduces the time to wait for data to be in place after each jump, and also reduces the delay (rotation delay) of the head waiting for the next sector to be in place on the same track. This accumulated tiny time savings are ultimately reflected in the "fluency" we feel. For example, I used to use a 5400 RPM laptop hard drive, and it may take a long time to open Photoshop. After changing to a 7200 RPM desktop hard drive, it feels like I have changed the world. Although it is incomparable to an SSD, the "stuttering" has indeed been greatly reduced. It makes every I/O operation respond faster, thus reducing the stuttering of applications as they wait for data.

Does the hard disk cache size really have such a big impact on performance?

The role of hard disk cache, or buffer, cannot be underestimated, but its "big impact" is often reflected in specific scenarios. The cache is like a small, high-speed temporary storage area inside the hard disk. When the system requests data, the hard disk will try to predict the data you may need next and load it into the cache in advance (read preview); or when you write data, the data will be written to the cache first, and the hard disk will then write to the slow disk at the right time (write later).

For those frequently accessed small files or data blocks that are read repeatedly, the advantages of large caches are very obvious. For example, some core files of the operating system, or index files in the database, are very likely to be read repeatedly. If this data can be kept in the cache for a long time, then each access can be directly retrieved from the cache, avoiding time-consuming mechanical seeking and reading. This can greatly improve random read and write performance, especially in multitasking environments where large caches can better smooth I/O peaks when multiple applications request hard disk data at the same time.

But then again, for files with very large single transfers, such as copying a movie with dozens of GB, the cache effect is relatively limited. Because no matter how large the cache is, it cannot install the entire file, and it ultimately depends on the physical read and write speed of the disk. It is more of a means of "assisted acceleration" and "optimized scheduling", rather than fundamentally changing the physical limits of mechanical hard disks. Therefore, if you mainly deal with sequential reading and writing of large files and the cache is increased from 64MB to 256MB, you may not feel a qualitative leap, but in daily operations and multitasking, the smoothness brought by large cache is real.

In addition to the hardware itself, what other software or system-level factors will slow down the mechanical hard drive?

In addition to the hardware specifications of the hard drive itself, many times, the hard drive slowness we feel is actually "stopping" at the system and software levels. The most common and most troublesome thing is undoubtedly the fragmentation of files . As mentioned earlier, when a file is written to the hard disk, if there is no continuous large enough space on the hard disk, the file will be divided into many small pieces, scattered in different locations of the disk. When we need to read this file, the head must be like a headless fly, "seeking treasures" everywhere on the disc, constantly moving and positioning, which greatly increases the time of seeking Tao, and naturally slows down. Regular disk defragmentation (although Windows will now automatically perform by default, it still needs to be paid attention to in old hard disks or specific usage scenarios) can effectively alleviate this problem.

Secondly, the operating system I/O scheduling strategy will also affect hard disk performance. The operating system will determine the execution order of hard disk requests based on different algorithms. For example, some schedulers will prioritize requests with high real-time requirements, while others will attempt to merge adjacent requests to reduce the number of seeks. Different strategies perform differently under different workloads, but usually modern operating systems are already doing well enough that we don’t need to intervene too much.

There are also some "invisible killers", such as applications running in the background . Antivirus software scans in the background, updates in the system indexing service, secretly downloading download tools, and even some rogue software frequently reads and writes log files. These will unconsciously occupy a large amount of hard disk I/O resources, resulting in the application you are using being slow to respond.

In addition, the health of the hard drive is also a key factor. Through SMART (Self-Monitoring, Analysis and Reporting Technology) data, we can see whether there are bad channels on the hard disk and whether the search error rate increases. A hard disk that is moving towards aging, wear of internal mechanical components will lead to increased seeking time, unstable data transmission, and even the risk of data loss. At this time, any software optimization is of no use, and replacing the hard drive is the only solution. The old or incompatible drivers, the efficiency differences in file systems (such as NTFS, ext4, etc.) in fragmentation management and metadata processing will also affect the actual performance of mechanical hard disks.

The above is the detailed content of What factors really determine the read and write speed of mechanical hard disks?. For more information, please follow other related articles on the PHP Chinese website!