Benchmarks
Unlike processors, most hard drive manufacturers do not bother with benchmark results. So, you will need to check hard drive reviews that actually do benchmark testing to find out about a particular hard drive's performance. To find out about your hard drive's performance, you can make use of these benchmarks.
HDTach 3 – The most basic of hard drive benchmarks, it focuses mostly on individual aspects of the drive's performance and will not give you a good idea of its real world capabilities. However, it offers a quick and easy way to check the drive's internal read/write speeds as well as its burst speed.
HDTune – Very much like HDTach and my comments are generally the same. It's fine as a basic benchmark, but hardware enthusiasts will want something with greater accuracy and consistency. On the other hand, it has a pretty good seek time benchmark which uses an averaged seek pattern generated from multiple seeks.
IOMeter – This is literally the God of server load testing. It started out as an Intel software but it is now managed and updated by the open source community. This benchmark offers a plethora of options and allows you to simulate random access with long queues and any other scenario imaginable. It is a top-notch benchmark. Most of the results come in IOPS (Input Output Per Second). However, you need to note that IOMeter is very server-centric and is a real pain to setup and use. That's why I treat my configuration files like gold.
IPEAK SPT – This is another legendary Intel benchmark. Like IOMeter, it has been discontinued by Intel and is no longer supported. However, this benchmark allows you to run anything you want on top of it. This is unlike IOMeter which generates its own load. With IPEAK, you set the load and IPEAK tells you how many IOPS the drive can push. Too bad it is so expensive at USD 800 and no longer under active development at that. A real shame.
PCMark – Well, the boys at Futuremark do make other software besides 3DMark, you know. PCMark is one of their offerings. Although I do think that their processor tests are awfully skewed in favour of Intel processors, their hard drive tests are pretty good with startup time tests based on real-world read/write loads. While it may not be the most accurate or detailed benchmark, they get a cookie for trying.
WinBench 99 V2.0 – Still one of the most highly-regarded test suites out there, although the company that produced it, Verisoft, was gobbled up by Lion's Gate and no longer supports it. You can still download a copy for your use though. The tests are very accurate and simulate real-world user loads extremely accurately. I certainly recommend you use this benchmark if you can.
Note : Remember that when you test your hard drives using these benchmarks, it is imperative that the hard drive's controller is not the bottleneck. Make sure your hard drive that supports SATA 3Gb/s is not set to run at only SATA 1.5Gb/s. Also, you should only compare results that were derived from equivalent hard drive controllers. There's no point in comparing hard drives tested with a low-end Silicon Image SATA controller with hard drives tested on a SAS controller with its own XOR processor from Promise or Dell, right?