In This Article
What is Disk Access Time?
Disk access time simply refers to the total time taken by the computer system as a whole to process a request for data sent by the processor and then retrieve that particular data from the storage device.
Technically, disk access time comprises the time taken to move the read and write heads to the chosen track, which is called the seek time, and the time taken to spin the disk so that the sector is found by the head, which is called the latency.
KEY TAKEAWAYS
- Disk access time is the time taken by the computer to process a data request and retrieve data from the different sectors of the hard disk where the data, which is particularly requested, is stored.
- The access time for the hard disk drive usually ranges within tens of milliseconds and is much higher than the access time of the SSDs, which is measured within the range of 25 to 100 microseconds.
- Usually, the two major components of the disk access time are the data access time and the data transfer time. However, the rotational latency of the disc as well as the command processing ability of the computer also play a significant role in it.
- The access time of the disk may vary depending on different parameters such as the data transfer rate or channel speed, the type of the drive and its performance, the interface quality, seek time and latency.
- You can reduce the disk access time by reducing the data transfer time. However, the disk access time is usually much higher than the seek time, which is different and a subpart of the disk access time.
Understanding Disk Access Time
Disk access time refers to the time taken by the computer to find the first data character once a data request is initiated.
In simple words, this access time of the hard disk involves the following:
- Initiation of a specific request made by the processor for a particular data
- Receipt of the initial bit of that particular data.
Typically, the access time taken by the disks is actually given as an average, always.
This is because it mainly depends on the seek time and rotational latency of the disk. Both of these parameters typically vary on the basis of the current location of the read and write heads on the platter.
Usually, the hard disk drives are known as direct access devices and they need different times to position the read and write heads of the disk, also called the head assembly or comb, on top of a particular data or track of the designated cylinder.
Apart from that, it also needs the disk to rotate so that the selected track of the particular record can be accessed by the comb.
The average disk access time is a very useful metric that helps in the proper analysis and planning of the computer application in real time, in an online inquiry system, for instance.
On the other hand, a minimum access time is very important for the proper and sequential use of disk drives.
Over a period of three decades or so, the rotational speed of the hard disk drives may not have improved significantly, with 3600 RPM being the standard, but the bit densities on each individual track have increased notably in the same period, by as much as fivefold.
This means that the average speeds for data transfer have also increased significantly. However, this has not reduced the access times for each track by a considerable margin.
Still, within this period, the average disk access times have been reduced by half. This is mainly due to the notable and widespread change in the design of the actuators in the hard disk drives.
Typically, in order to move the head assembly mechanism, the hard disk drives now come with voice coil actuators in place of the traditional hydraulic actuators.
Variance
The disk access times typically differ depending on the type of the drive. For example, the different types of Hard Disk Drives, the respective access times for each will be as follows:
- For enterprise quality HDDs, such as Scorpio Black and WD (Veloci) Raptor, the average disk access time will be less than 10 ms.
- For desktop quality HDDs, the average disk access time will range anywhere between 12 ms and 15 ms.
- For the less-used floppy disk drives, the average access time is about 25 ms.
- For the Compact Disc Read Only Memories (CD-ROMs), the disk access time is typically about double that of the floppy disks.
- For the videotape cartridges with honeycomb lattices and other similar types of mass storage systems, such as the tape-cartridge systems, the average access time ranges between 12 and 15 seconds.
However, for the Dynamic Random Access Memory or DRAM-based storage drives, the average disk access time will not be enough to measure.
And as for the Solid State Drives, just as their name implies, they do not have the usual components of an HDD.
Therefore, in the SSDs, the delay in getting the data will be very short and may be measured in nanoseconds.
The actual rate of reading data from the storage array will also be pretty short. So, the disk access time for an SSD will be much less in comparison to the disk access time of an HDD.
However, the probable limitation will be in the capabilities and speed of the other components within the SSD as well as the interface connecting it to the rest of the computer system in terms of its speed and bandwidth.
As a result, for the SSDs, the average disk access time will be in the tens of microseconds.
Disk Access Time Formula
The disk access time can be calculated easily by using a simple formula: Disk Access Time = Access Time + Data Transfer Time.
However, there are other factors to consider to calculate the actual disk access time, and when these are considered, the formula would be:
Disk Access Time = Seek time + Rotational delay + Transfer time + Controller overhead + Queuing delay.
Here, the different elements of the equation have to be calculated differently, as explained hereunder.
Average seek time is calculated as follows:
Average seek time = 1/3 x time taken for one full stroke.
For example, if the time taken by the read and write head to move from one track to the next track is t units and the number of tracks in total is k, then the average seek time will be:
{Time taken to move from track 1 to track 1 + Time taken to move from track 1 to last track} / 2
= {0 + (k-1) t} / 2
= (k-1) t / 2.
The average rotational latency is calculated as = 1 / 2 x time taken for one full rotation.
The capacity of a disk pack is calculated as = Total number of surfaces x Number of tracks per surface x Number of sectors per track x Storage capacity of one sector.
Formatting overhead is calculated as = Number of sectors x Overhead per sector.
Formatted disk space or usable disk space is calculated as = Total disk space or capacity – Formatting overhead.
Recording density or storage density of a track is calculated as = Capacity of the track / Circumference of the track.
Capacity of a track is calculated as = Recording density of the track x Circumference of the track.
Data transfer rate is calculated as = Number of heads x Bytes that can be read in one full rotation x Number of rotations in one second.
It can also be calculated alternatively as = Number of heads x Capacity of one track x Number of rotations in one second.
Total number of tracks on each surface is calculated as = (Outer radius – Inner radius) / Inter track gap.
The controller time is equal to the overhead levied by the disk controller.
Disk Access Time Example
The disk access time can range between 12 milliseconds to 20 milliseconds usually, depending on the parameters and type of hard disks. However, it will be different for the enterprise drives, desktop drives, and the Solid State Drives (SSDs).
For example, assume that a hard disk comes with the following parameters:
- Transfer size of 8K bytes
- Advertised average seek of 12ms
- Disk rotational speed of 7200 RPM
- Data transfer rate of 4 MB/sec
- Controller overhead of 2ms
Further, assume that there is no queuing delay since the disk of the drive is supposed to be idle and the advertised seek time undertakes no locality.
With all these things considered, the average disk access time for a given sector on it will be as follows:
12 ms + 0.5/ (7200 RPM/60) + 8 KB/4 MB/s + 2 ms
= 12 + 4.15 + 2 + 2
= 20 ms
Now, this access time of the disks may also vary depending on the type of the hard drive in the first place, as said earlier.
What are the Components of Access Time in a Disk?
The main components of disk access time are the seek time, data transfer time, processing time of the commands, and rotational latency. All these components are considered and added together to obtain the access time of a disk.
Seek time:
The seek time is the time taken to move the read and write heads from one location to another location on the disk to read data or locate the desired track. The higher the seek time is, the greater the data access time will be.
Rotational latency:
Rotational latency or wait time refers to the time the read and write head has to wait for the disk to spin around so that the desired sector comes under it and it can read the data stored on that particular sector.
Data transfer speed:
The data transfer speed typically involves two separate components. These are:
- The speed at which the disk rotates and the bits pass under the read and write heads
- The speed of the interface connecting the hard disk to the rest of the computer
The slower the speed, the higher will be the disk access time, especially for making large data block transfers.
Command processing time:
And the processing time taken by the computer to deduce the commands and instructions also adds to the total disk access time and data transfer. Though it mainly depends on the Central Processing Unit, all other components of the computer also play a significant role in it.
What is the Average Disk Access Time for a Sector?
Typically, in most cases, the average disk access time is about 12 milliseconds to 20 milliseconds.
However, it may vary according to the type of the hard drive as well as other parameters such as:
- The seek time
- The sector read time
- The average rotational delay
- The overhead.
In addition to that, it is also quite likely that the working process of the disk controller will affect the disk access time. This is because it may need the controller to perform two separate operations in order to read a sector such as:
- Read into the hardware buffer first, which will take a few microseconds to complete and
- Transfer the data to the memory after that, adding to the time.
In such cases, the average disk access time will be higher, even higher than 20 ms.
However, at this point it is good to keep in mind that the type and working process of the hard disk and its controller will affect the data access times quite significantly.
Disk Access Time vs Seek Time
- Disk access time refers to the time required by the computer system to process all read and write requests and salvage the necessary data. On the other hand, seek time refers to the time needed by the read and write heads to move from one location on the disk to another sector.
- The disk access time is quite high in comparison to the seek time because the latter is a subpart of the former.
- In calculating the disk access time, the time taken to transfer data is also considered. However, it is not required while calculating the seek time.
- Disk access time primarily depends on two specific parts, namely the access time and data transfer time. On the other hand, the seek time can vary a lot because it depends on the distance to be travelled by the read and write head between the current and final positions and how it is instructed to move.
- The disk access time can be reduced only if the data transfer time and access time are reduced. On the other hand, the seek time usually stays between 10 ms and 20 ms, depending on the drive grade and the RPS.
- There are different parameters that determine the disk access time such as the seek time, the rotational latency, and the data transfer time. However, in comparison, the seek time is measured on the average time taken to find something.
Conclusion
Just as the article points out, disk access time is the time taken by the computer to process and retrieve the results of the read and write requests.
It is usually measured in milliseconds but is relatively higher than data seek time, the two being completely different things, as this article has also pointed out.