Today all leading companies have hard drives with the buffer size increased up to 8 MB. In Western Digital JB series the scores got better in Ziff-Davis WinMarks but the performance worsened in all access models of Intel IOMeter. Now it's Maxtor's turn.
Model 6Y080P0 of 80GB is included into DiamondMax Plus9 line described in the review of models '02. This is a full-size line of the density of 80GB per platter (according to the manufacturer). But our watchful readers noticed that the platter capacity could vary. The discs of firmware version YAR41VW0 feature 60 GB per platter (80 GB is obtainable with more tracks), the discs of firmware version YAR41BW0 have it dependent on a disc size - 80 GB per platter for 80GB and 160GB drives and 60GB per platter for 60GB and 120GB drives. I haven't tested yet drives with the newer firmware version, but I will do it soon. The discs can have 60, 80, 120 and 160 GB (200GB drive have recently appeared as well). The codename is 6Yxxxz0, where xxx is capacity, z is a disc type (L - standard with a 2MB buffer, P - with a 8MB buffer, M - with Serial ATA interface).
The 2MB and 8MB drives differ only in the cache memory chip (visually).
The full drives' specs are given in the table.
The test system is standard.
The test suite is standard as well.
For comparison we used results of 60GB model 6Y060L0 with a 2MB buffer. The comparison is not entirely correct, that is why model 6Y080L0 will be thoroughly tested in the near future.
The drives were tested in 4 modes - with factory settings (acoustic level = fast) and with acoustic management off with the drives connected either to ATA/133 controller or to built-in ATA/100 controller. The acoustic management status hardly influenced the finals, that is why further tests were carried out with acoustic management off
Ziff-Davis WinBench 99 / HDTach 2.61
Model 6Y080P0 lost because of the strange linear reading in the beginning. Is it because of the buffer? Maybe.
Model 6Y080P0 keeps ahead at the average speeds. The scores are given for ATA/133 mode.
Fortunately, the access time got lower. If you remember, Western Digital drives of 8MB buffer had it worse than the standard models.
The performance grows up thanks to the faster interface and the larger buffer.
Intel IOMeter - 7200 rpm
In case of low and average loads the drive with the larger buffer wins. In case of larger loads the scores depend on the access model. On the whole, the scores got better with the larger buffer, but the ATA/133 interface brings no benefit.
These diagrams explain the situation. In contrast to WD drives, here the 8MB buffer boosts the speed in case of random writing. In case of random reading it wins only with low loads - a big buffer can't be an advantage with large loads, i.e. a large number of simultaneous requests, because reading is random.
Finally, let's compare the drive with the other models.
In my opinion, Maxtor 6Y080P0 is a drive with the well debugged firmware
optimized for real applications (or tests based on them :)). It performs
worse in the synthetic tests, but anyway, the model of the 8MB buffer outdoes
the standard solution.
Nikolai Dorofeev (firstname.lastname@example.org)
Write a comment below. No registration needed!
Copyright © Byrds Research & Publishing, Ltd., 1997–2011. All rights reserved.