Hard drives - 500GB

While a good hard disk isn't likely to elicit the same 'wow!' response that a new CPU or graphics card will generate, without a hard disk drive, your PC would be nothing more than a stupidly fast pocket calculator. A hard disk drive is an essential component; after all, before hard disk drives, PCs had to boot DOS from one floppy disk, and then load apps or games from another. Disk drives also affect your PC's performance, determining how long apps and games take to load, and how quickly your PC boots up.

Choosing a hard disk is much like choosing memory. First, you decide what capacity you want/need, and then you select the best model. However, this isn't nearly as easy as it looks, despite hard disks being marketed as members of a series. This is because, despite often having similar model names, there's frequently little commonality between two drives from the same series. For example, the Hitachi Deskstar T7K500 series comprises 12 models, which span two different interfaces (EIDE and S-ATA II), four different capacities (250, 320, 400 and 500GB) and two buffer sizes (8MB and 16MB). To further complicate matters, drives with different capacities also have different areal and track densities, which have a massive effect on performance.

We've rounded up the major models of every current generation of S-ATA II hard disk, ranging from a minuscule 74GB drive to a titanic 750GB model. However, instead of writing individual reviews of each disk (which would quickly become extremely repetitive), we've arranged the reviews by capacity, so you can be sure of buying the best drive for your PC.

We perform two broad types of performance testing on hard disk drives. The first tests use Simpli Software's HD Tach 3 RW (www.simplisoftware.com) application, which runs a series of benchmarks to record sustained transfer rate (STR) when reading and writing, plus average access time. Testing is performed across the full surface of all the disk's platters to obtain the lowest and highest transfer rates, as well as the average rates.

Although it's a synthetic benchmark, the information from the HD Tach 3 RW test helps to explain real-world performance, and raw throughput is fundamentally important for tasks such as video editing. However, a hard disk's performance in real-world applications is affected by other factors too, such as the memory buffer size, and the algorithms used to calculate which data to store in the buffer and which to discard. For this reason, we also test hard disks using several real-world scenarios, which are based on the kinds of tasks that your PC is likely to perform.

Our first real-world test is the Paint Shop Pro image editing test from our Media Benchmarks 2006. This test continually reads and writes data to and from the hard disk, which means that a slow drive can have a significant effect on performance.

Our second real-world test times how long the test PC takes to load the Fort level of Far Cry. At first glance, this looks like the kind of task that would benefit from a drive with a fast sustained throughput, because games tend to store much of their content in large archive files. However, in reality, most game level loading involves grabbing smaller files from within these archives, which means that access time is just as important.

Our current hard disk test rig consists of a 2.13GHz Core 2 Quad Q6400 CPU installed in an Asus P5B-E Plus motherboard, with 1GB of Kingston PC2-6400 RAM. This is typical of the kind of performance rig to which most of us aspire, as the motherboard is based on the excellent Intel P965 chipset with Intel's ICH8R Southbridge. The ICH8R Southbridge is particularly important, since it supports both S-ATA II and NCQ.

The overall Speed score is based on a weighted combination of the different real-world and synthetic performance tests. We don't include a separate Features score for hard disk drives, as most features directly relate to performance (buffer size, areal density and so on), although factors such as low noise levels (or a viewing window) add value, and therefore slightly increase the Value score. However, the bulk of the Value score is based on a weighted calculation of performance and price per GB. The Overall score is the sum of these two scores.