How to Select a Hard Drive Based on Specifications

Hard drive specifications for computers are generally the easiest to understand. There are really only two numbers that are needed to know: size and speed.

All hard drive manufacturers and computer systems rate their drives in GB or gigabytes. This translates to the unformatted capacity of the drive in billion of bytes. Once the drive is formatted, you will actually have slightly less than this number in drive space. This makes size comparison really easy to determine as the higher the number, the larger the drive.

Most consumer desktop systems come with either a 5400 rpm or 7200 rpm hard drive. Systems which use the 5400 rpm drives tend to be slower but are used in more compact systems or systems which are designed to be silent running. It is more common to see drives running at the 7200 rpm speed as they give better performance. Now wasn't that simple?

So What to Get?

Determining what type of hard drive you should get for in your computer depends really upon what type of tasks you will be using the computer for. Different tasks require various sizes of file storage as well as performance. Below is a chart that lists some of the common computing tasks relating to what size and speed hard drive to look for in a system:

  • Word Processing: 40+ GB, 5400 rpm
  • Web Surfing: 40+ GB, 5400 rpm
  • Light Gaming: 80+ GB. 7200 rpm
  • Digital Music: 120+ GB, 7200 rpm
  • Heavy Gaming: 100+ GB, 7200 rpm
  • Graphics Editing: 160+ GB, 7200 rpm
  • Digital Video: 200+ GB, 7200 rpm

These are just general guidelines considering the most common amounts of storage space that files and programs associated with these tasks cost. With the current size and cost of hard drives for computer systems, it is easy to find drives of larger capacity than the numbers listed above for very little in cost.

IDE and Serial ATA:

Recently a new type of drive interface has made its way into the marketplace. Not all computer manufacturers will list the type of interface used with the hard drive. For most people, the differences between the two are very minimal. The performance between the two interfaces is essentially identical at this point. The major difference really is the ease of installing the drives. Serial ATA drives have less cabling and configuration required to install a drive. More information on the new interface can be found in the article.

Hard drive and optical storage devices for the personal computer market have been using variations of the IDE and ATA standards for over 10 years now. It has undergone a great many revisions over the years and is finally reaching the limits of its capabilities. To address the limitations of the ATA interface, the new Serial ATA interface has been under development for several years and is now finally starting to become available to consumers in the forms of controllers and more importantly hard drives. So lets take a look at this new interface and what its capabilities are compared to the older ATA format.

Performance:

Performance is the driving factor for most people's hard drive selection. A slow hard drive directly impacts all of your computing tasks. Hard drive performance is really determined by four core attributes of a drive:

  1. Interface
  2. Rotational Speed
  3. Access Times
  4. Buffer Size

There are currently two primary interfaces used for hard drives for personal computers on the market: SCSI and IDE. SCSI is used for high performance drive arrays and server systems and generally not found in the home computer market. Its advantage is its ability to handle many drives concurrently. The current SCSI standard is Ultra 160 that allows for 160 megabytes of data per second.

IDE interfaces are the most common form of interface found on personal computers. There are a number of speeds available for IDE ranging from ATA/33 to ATA/133. Most drives support up to the ATA/100 standard and are backward compatible with older versions. The number in the version indicates the maximum bandwidth in megabytes per second the interface can handle. Thus, an ATA/100 interface can support 100 MB/sec. Currently no hard drive is able to reach these sustained transfer rates, so anything beyond ATA/100 is not needed. Also, systems are restricted to the PCI Bus speed that can only handle about 133 megabytes per second.

The biggest drawback to the IDE standard is how it handles multiple devices. Each IDE controller has 2 channels that in turn can support 2 devices. The controller must however scale its speed to the slowest device on the channel. This is why you see 2 IDE channels: one for hard drives and a second for optical drives. A hard drive and optical drive on the same channel results in the controller scaling back its performance to the optical drive speed which degrades performance for the hard drive.

The rotational speed of the disks in the drives is the largest factor in the performance of the drive. The higher the rotational speed of the drive, the more data the drive can read and write from the drive in a fixed amount of time. Heat and noise are the two byproducts of higher rotational speed. Heat impacts the performance of the electronics within the computer, especially if there is poor ventilation. Noise can cause distractions for people in or around the computer. Most home computer hard drives rotate at either 5400 or 7200 rpm. Some higher speed server drives run at 10,000 rpm or higher.

Access times refer to the length of time it takes the drive to position the drive head on the platter for the appropriate function. There are generally 4 access times listed for all hard drives on the market:

  • Read Seek
  • Write Seek
  • Track-to-Track
  • Full Stroke

All 4 of the items are rated in milliseconds. Read seek is generally an average time it takes to move the head from one position on the drive to another to read data from the drive. Write seek is the average amount of time that it takes the drive to move to an empty space on the disk and begin writing the data. Track-to-track is the average amount of time the drive takes to move the drive head to each sequential track on the drive. Full stroke is the amount of time it takes the drive head to move from the outer to inner portion of the disk or the full length of the drive head's motion. For all of these, a lower number means higher performance.

The final factor that impacts performance for a hard drive is the amount of buffer on the drive. A drive's buffer is an amount of RAM on the drive to store frequently accessed data from the drive. Since RAM is faster at transferring data than the drive head operation, it increases the speed of the drive. The more buffer on the drive, the more data that can be stored in the cache to decrease the amount of physical drive operation. Most drives today come with a 2MB drive buffer. Some performance drives such as the Western Digital Caviar Special Edition drives ship with a 8MB buffer.

Capacity:

Hard drive capacity is the most simplistic of the factors when looking at a hard drive and also what most companies will try to stress to consumers. Every drive on the market is marketed at a specific storage capacity in gigabyte (GB). One gigabyte is actually 1 billion bytes rather than the literal 1,073,741,824 bytes. What people don't always realize is how the capacity on the actual platters can translate into performance and aesthetics.

Every drive is made of one or more disk platters each with its own drive heads to read and write the information from the platters. Each of the platters has a rated capacity. For example, each platter may be 20 GB. A manufacture can then produce 3 drives all from the same platter by varying the number of platters in the drive to produce 20, 40 or 60 GB drives.

Additional platters can increase performance as the number of drive heads to read the data is increased but this also generally translates into a higher ambient noise from the additional platters and drive heads. Also, the more drive platters there are, the more energy that must be expended to turn the platters at the proper speeds. This increases the amount of heat produced by the drive that can actually decrease the overall performance compared to a similar capacity drive with fewer platters. So, more platters does not always mean a better performing hard drive.

Aesthetics:

Many people don't realize there are aesthetics to a component that resides inside of a computer. Many manufacturers refer to these as the environmental specifications for their drives. All drives produce two byproducts to their operation just like all other computer equipment, noise and heat.

Anyone who has ever heard an original Apple II floppy drive can attest to the noise a drive can produce. That drive sounded like it was grinding up any floppy that was inserted into it. Now the computer case may dampen the noise, but a noisy drive can be a distraction for some people. In this case, they may sacrifice some performance for a more tolerable computer environment. In general, slower rotational drives tend to produce less noise than higher rotational speed drives. All ratings for a drives noise are listed in decibels (dBA). The higher the number of decibels, the louder the drive is.

The heat byproduct affects aesthetics but can also impact the lifespan of your computer system. All drives generate heat from their operation. Prolonged operation can even cause premature failure of the hard drive. Some of the more recent IBM Deskstar drives are only rated for periodic usage. Prolonged usage of the drives caused the drives to fail due to the excessive heat generated by the drive. This is amplified if the computer case it resides in has poor ventilation to dissipate the heat.

Excess heat can also be detrimental to the user. Computer systems that produce a lot of heat tend to heat their environment. In cold climates this may not be much of a problem, but in warmer climates it can once again reduce the life span of the drive and make it unpleasant to be in the same room. Hard drives are not rated on how much heat the drives produce as a by-product. This makes it near impossible to compare or even shop for in a hard drive.

The Proper Balance:

So, what does make the right drive for your computer? The answer depends on the all around desires for the user. With the proper system configuration and if you aren't too concerned about noise or heat, then a high performance drive might be the answer. Maybe having the largest storage space is the desire. For some it might be that they can install the drive into their computer that won't disrupt them while they work or turn the room into a furnace.

When shopping for a hard drive, here are some of the numbers that are published by the manufacturers that you should look at:

  • Capacity - in gigabytes
  • Rotational Speed - rotations per minute
  • Read Seek (average) - milliseconds
  • Write Seek (average) - milliseconds
  • Track-to-Track (average) - milliseconds
  • Full Stroke (average) - milliseconds
  • Idle Noise - in decibels (dBA)
  • Seek Noise - in decibels (dBA)

Bits, Bytes and Prefixes:

All computer data is stored in a binary format as either a one or zero. Eight of these bits together for the most commonly referred to item in computing, the byte. The various amounts of storage capacity are referred by a prefix to represent a specific amount, similar to the metric prefixes.

Since all computers are based on binary math, these prefixes represent base 2 amounts. Each level is an increment of 2 to the 10th power or 1,024. The common prefixes are as follows:

  • Kilobyte (KB) = 1,024 Bytes
  • MegaByte (MB) = 1,024 Kilobytes or 1,048,576 Bytes
  • Gigabyte (GB) = 1,024 Megabytes or 1,073,741,824 Bytes
  • Terabyte (TB) = 1,024 Gigabytes or 1,099,511,627,776 Bytes

This is very important information because when a computer operating system or program reports the available space on a drive, it is going to report the overall total of available bytes or reference them by one of the prefixes. So, an OS reporting a total space of 70.4 GB actually has around 75,591,424,409 Bytes of storage space.

Advertised vs. Actual:

Since consumers don't think in base 2 mathematics, manufacturers decided to rate most drive capacities based on the standard base 10 numbers we are all familiar with. Therefore, one Megabyte equals one million bytes while one Gigabyte equals one billion bytes. This isn't too much of a problem with fairly small numbers such as a Kilobyte, but each level of increase in the prefix also increased the total discrepancy of the actual space compared to the advertised space.

Here is a quick reference to show the amount that the actual values differ compared to the advertised for each common referenced value:

  • Megabyte Difference = 48,576 Bytes
  • Gigabyte Difference = 73,741,824 Bytes
  • Terabyte Difference = 99,511,627,776 Bytes

Based on this, for each Gigabyte that a drive manufacturer claims, they are over reporting the amount of disk space by 73,741,824 Bytes or roughly 70.3 MB of disk space. So, if a manufacturer advertises an 80 GB (80 billion bytes) hard drive, the actual disk space is around 74.5 GB of space, roughly 7% less than what they advertise.

Now, this isn't true for all the drives and storage media on the market. This is where consumers have to be careful. Most hard drives are reported based on the advertised values where a Gigabyte is one billion bytes. On the other hand, most flash media storage is based around the actual memory amounts. So a 512MB memory card has exactly 512 MB of data capacity, but this leads to the next area of reported space.

Formatted vs. Unformatted:

In order for any type of storage device to be functional, there must be some method for the computer to know which bits stored on it relate to the specific files. This is where formatting of a drive comes in. The types of drive formats can vary depending on the computer but some of the more common ones are FAT16, FAT32 and NTFS. In each of these formatting schemes, a portion of the storage space is allocated so that the data on the drive can be catalogued enabling the computer or other device to properly read and write the data to the drive.

This means that when a drive is formatted, the functional storage space of the drive will be less than its unformatted capacity. The amount by which the space is reduced will vary depending upon the type of formatting used for the drive and also the amount and size of the various files on the system. Since it does vary, it is impossible for the manufacturers to quote the formatted size. This problem is most frequently encountered with flash media storage over larger capacity hard drives.