Articles: Storage
 

Bookmark and Share

(2) 
Pages: [ 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 ]

Performance in IOMark

We use our internal IOMark tool for low-level tests. Let’s check out the sequential read speed of the drives first.

Now we can compare the HDDs by the read speed at the beginning and end of the partitions created on them.

We’ve got a very interesting picture here. While it is all clear with the 2TB Western Digital (the 5400rpm drive is as fast as the previous-generation 7200rpm model thanks to increased recording density), things are more complicated with Seagate’s products. The highest speed at the beginning of the partition is delivered by the 1.5TB Barracuda 7200.11 rather than by the Barracuda 7200.12 but the new 2-platter 1TB model has the highest speed at the end of the partition. It looks like Seagate has been unable to ensure correct operation of the heads over the area where sectors run under the head in the shortest time (these are the outermost sectors, as you can guess) and agreed to a certain reduction of recording density in it. And there is one more thing to be noted: Seagate’s drives draw rather smooth data-transfer graphs whereas the new drive from Western Digital has very wild fluctuations of speed at the beginning of the partition. Yes, achieving such a high recording density has not been easy for anyone. With Western Digital, the speed depends heavily on the lucky combination of platter surface and head.

And what about cache memory speed?

The new Seagate Barracuda 7200.12 seems to have new electronics that helps it beat its opponents, including the Barracuda 7200.11 series products. Judging by the shape of the graphs, Seagate’s programmers are working hard on the firmware: the 1.5TB Barracuda 7200.11 model is indicative of their attempt to solve the problem of low performance with data blocks larger than 256 sectors (128 KB). The attempt was not quite successful, though. With the Barracuda 7200.12 the problem is nearly solved for reading: there are but rare slumps instead of peaks on a flat plane. However, the company still has not got rid of the slump of the buffer write speed when the HDD processes large data blocks.

The 2TB drive from Western Digital does not show anything new. The performance hit when writing in large data blocks is not as serious as to require immediate reaction. The developer just wisely follows the old rule: do not repair things that already work!

 
Pages: [ 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 ]

Discussion

Comments currently: 2
Discussion started: 05/10/09 05:37:36 AM
Latest comment: 11/10/10 04:29:28 AM

View comments

Add your Comment