Since our last hard drive report published in August 2010 and updated earlier this year, several manufacturers have introduced new models to their ranges. Hitachi have brought in the Deskstar 5K3000 and 7K3000 and Seagate's 5900 rpm range (previously Barracuda LPs) are now listed as ‘Green’.
7 models, but 8 disk drives
We tested the following HDDs for this comparative:
From Hitachi, we have the Deskstar 7K3000 (HDS723020BLA642) and the Deskstar 5K3000 (HDS5C3020ALA632). These two drives run at 7200 and 5700 rpm and have 64 and 32 MB of cache. They both have an SATA 6G interface and three platters each with a density of 667 GB.
Samsung have one drive only at this capacity, the EcoGreen F4 (H204UI). With an SATA 3G interface, which isn’t problematic given that hard drives are far from saturating it, it has 32 MB of cache and is equipped with three 667 GB platters.
From Seagate, we have the Barracuda XT (ST32000641AS) and Barracuda Green (ST2000DL003). Here both drives have an SATA 6G interface and a 64 MB cache. They run at 7200 and 5900 rpm respectively and while the XT has four 500 GB platters, the Green has three platters at 667 GB.
We also tested three drives from Western Digital. The WD Caviar Black (WD2002FAEX) has an SATA 3G interface and 64 MB of cache. It has four 500 GB platters with a rotation speed of 7200 rpm. There are two versions of the WD Caviar Green WD20EARS: the WD20EARS-00MVWB0 and WD20EARS-00J2GB0, which have three 667 GB and four 500 GB platters respectively.
Though the second of these models is less widely available, it is still found on sale and we decided to test it in view of the fact that Western hasn’t bothered to give it a separate reference. These drives have an SATA 3G interface and a 64 MB cache. A new Caviar Green model has also recently appeared, the WD20EARX. According to Western’s website, it only differs by its SATA 6G interface and it’s to be hoped that this means it will use three platters.
For this report we used the test protocol we put into place for our 2011 SSD comparative, though with a couple of changes. We used h2bench for sequential speeds however, so as to take our readings according to where the read head is positioned on the drive. In addition we carried out fewer practical tests, limiting ourselves to launching 3D Studio Max, launching a level in Crysis and installing Photoshop CS 5. All the tests were carried out on a P67 with an SATA 6G interface. The drives are however far from saturating this interface and moving over to 3G has virtually no impact on the performance of models that support 6G.
Of course we added specific hard drive tests, namely temperature and noise level measurements. Noise levels were measured with a Cirrus Optimus CR152A Class 2 sonometer, which can measure levels as low as 21 dBA, the lowest level valid in the room used for testing. This is the second time we have used this high-end sonometer, having first used it in our cooler comparative