Articles: Storage
 

Bookmark and Share

(3) 
Pages: [ 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 ]

Testbed and Methods

The following testing utilities were used:

  • IOMark 0.30b14
  • IOMeter 2003.02.15
  • FC-Test 1.0
  • PCMark Vantage
  • Windows 7 Disk Defragmenter
  • WinRAR 3.91

Testbed configuration:

  • ASUS P5WDG2 WS Pro mainboard
  • Intel Pentium 4 620 processor
  • IBM DTLA-307015 system disk, 15 GB
  • Radeon X600 graphics card
  • 1 GB DDR2-800 SDRAM
  • Microsoft Windows 7 Ultimate

HDDs are tested with the generic OS drivers. We format them in NTFS as one partition with the default cluster size (for FC-Test we create 32GB partitions), connect them to a mainboard port and enable AHCI. We have transitioned to new hard disk drives testing methodology.

As this is the first time we see a 4KB-sector drive from Samsung, we will test the EcoGreen F4 in two modes: aligned and not aligned. For the latter mode we will format the partition created on this HDD in Windows XP even though our methodology implies the use of Windows 7. We will also use IOMeter loads which are not aligned with the 4KB sectors.

Performance in IOMark

We use our internal IOMark tool for low-level tests. Let’s begin with sequential reading.

Let’s compare the drives according to the read speed at the beginning and end of the full-capacity partitions created on them.

This test becomes more and more inadequate, so we think about abandoning it altogether. Everything depends on whether the particular HDD is lucky to have a fast head/surface pair in the first sectors. As a result, the comparison in terms of bottom and top speeds doesn’t produce correct results anymore. As a matter of fact, there is no big difference between the HDDs with 500GB platters whereas the SpinPoint F4 is somewhat faster than them. And the new WD Caviar Green is actually no slower than its energy-efficient opponents. We can see all this clearly in the graphs but the summary diagram gives us no clue about that. On the other hand, comparing over a dozen of graphs with each other is quite a daunting task and rather useless too as we will be able to see the HDDs’ top speeds in IOMeter’s sequential read test.

Still, the graphs deserve a closer look. For example, they make it clear than the 1.5-terabyte drives are not cut-down versions of the 2-terabyte models. In both cases we are dealing with full-featured 3-platter products.

Now, what about the speed of working with the cache buffer?

Everything is simple here: the Hitachi drives are ahead in terms of top read speed and the Samsung team wins at writing. Samsung’s HDDs produce ideal-looking graphs. As opposed to them the Hitachi drives show their typical dependence of burst speed on the size of the data block. Their graphs have peaks and slumps, the new firmware changing nothing in this respect. As for the drives from WD, they are good as always in this test. The 1-terabyte models and the new Caviar Green have even got rid of the performance hit on large data blocks which could be observed with the older drives from WD.

 
Pages: [ 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 ]

Discussion

Comments currently: 3
Discussion started: 12/29/10 06:18:57 AM
Latest comment: 01/09/11 08:57:17 PM

View comments

Add your Comment