Articles: Storage

Bookmark and Share

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.

Judging by the graphs, the manufacturers all try to squeeze everything out of the drives: the speed is considerably higher with lucky surface/head combinations and the graphs are not smooth but fluctuating. The resulting diagram does not represent the situation correctly because the numbers depends greatly on whether the HDD has a lucky surface/head combination at the beginning and end. The only thing we are sure of is that the Seagate is faster than the others. The Caviar Green is expectedly slower than the HDDs with higher spindle rotation speed, the difference amounting to about 30MBps.

And what about buffer memory? We will check this out using a test that reads and writes data in small blocks that fit entirely into the cache buffer.


The HDDs are all about the same with data blocks up to 128KB, delivering speeds above 200MBps at both reading and writing, the Samsung and Seagate being somewhat faster than the drives from Western Digital. The picture is more interesting when you take a look at the graphs. When the data block is larger than 128KB (256 sectors), Western Digital’s drives have a temporary reduction of speed, but then their burst read speed begins to increase. The speed of writing doesn’t increase and shows rather odd fluctuations. The Samsung is neater and smoother: it suffers the same performance hit and delivers normal speed of writing. The Seagate has uniform speed but its performance hit is heavier: the burst read speed returns to the level of 170MBps while the burst write speed is barely above 100MBps when the drive is processing large data blocks.

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


Comments currently: 11
Discussion started: 03/08/09 06:25:48 AM
Latest comment: 12/21/15 11:36:13 AM

View comments

Add your Comment