Product survey: 5 SATA 500 GB HDD - BeHardware
>> Storage >> Hard Disks

Written by Marc Prieur

Published on August 14, 2007

URL: http://www.behardware.com/art/lire/679/


Page 1

Introduction



In our first hard drive survey a year ago, we had to plan on spending almost 300 Ä for a storage capacity of 500 GB. This price has since been decreased by two-thirds and now itís no longer rare to find such a capacity for 100 Ä or less. Seagateís acquisition of Maxtor didnít slow the competition down and Samsung even gained some power in the PC hard drive market.

5, 4, 3 and soon 2 platters
A year ago, only Seagate relied on perpendicular recording for this type of model allowing them to produce 500 GB hard drives with only three platters where Maxtor and Western used four and Hitachi five. As for Samsung, they were simply absent from this market.

You may recall, perpendicular recording allows the reduction of the surface occupied by each bit of data resulting in higher surface density. Therefore, more Gigabytes can be put on the same platter. In todayís survey, four of the five products have a three platter design, the Hitachi T7K500, Samsung T166 (HD501LJ), Seagate 7200.10 as well as the Western Caviar SE16 (WD5000AAKS). As a reference, weíve also included a Hitachi E7K500, formerly known as the 7K500, which is equipped with no less than five platters. All of these hard drives have a cache of 16 MB.


We should note that in the last few weeks several announcements have appeared foreshadowing upcoming innovations in this market. Seagate announced its 7200.11, while Samsung is launching its SpinPoint F1. Each formerly had 167 GB per platter and we now make the transition to 250 and 333 GB, respectively, allowing these two manufacturers to make 500 GB hard drives with two platters. For Samsung there will be single sided use. This should allow price reductions to continue with 500 GB priced at around 70 Ä next year. The only thing is that announcements arenít synonymous with availability and we will have to wait a little before seeing these products in stores.


Page 2
The hard drives, the test

The hard drives in pictures
Here are the hard drives in this test with the firmware reference in parenthesis:

- Hitachi E7K500 (K2AOAD1A)
- Hitachi T7K500 (V56OA7EA)
- Samsung T166 - HD501LJ (CR100-10)
- Seagate Barracuda 7200.10 (3.AAK)
- Western Digital Caviar SE16 Ė WD5000AAKS (12.01C01)

Hitachi E7K500Hitachi T7K500
Samsung T166Seagate 7200.10
Western Caviar SE16

There is nothing particularly noteworthy on the front side.

Hitachi E7K500Hitachi T7K500
Samsung T166Seagate 7200.10
Western Caviar SE16

On the back, we can see that each brand uses a different controller. While Hitachi is true to Infineon and Seagate to Agere, Samsung and Western both go with Marvell, however, not the same model and we donít know the exact differences.
Test protocol
We took numerous measurements in this test. First of all, we were interested in ę synthetic Ľ performances, or in other words, cache rates, sequential speed, and average access time. Next came application tests in the form of an application performance index based on PCMark05, a file server load type simulation via IOMeter, and finally the writing, reading and close (on the same partition) and far (on the partition that starts at 50% of the hard drive) copying a collection of files.

These consist of two big files for a total of 4.4 GB plus 2620 files amounting to 2 GB and finally 16046 files for a total of 733 MB. The source when reading or writing was a RAID of two 74 GB Raptors capable of assuring a rate of 110 MB /s in order to not be limited on this side. This test is of course interesting because while sequential rates give us an idea of performances when copying big files, things can be different with a series of smaller ones.

All measurements were taken with sound control deactivated. We took some measurements afterwards with this function (except for the Seagate because it isnít available). The test machine was based on a P965 Express chipset mounted on an ASUSTeK P5B Deluxe motherboard and Serial ATA ports were configured to AHCI (Advanced Host Controller Interface) in the bios in order to use NCQ.

Of course, in addition to performance measurements we took a look at the temperature after two hours of intensive use as well as power consumption. Finally, noise levels were evaluated via dBA measurements and we also give you an actual recording of each hard drive in various states.


Page 3
Synthetic performances

Synthetic performances


Here are the rates obtained in the ęCore TestĽ of h2bench which allows measuring cache speed. You may recall that the cache is 16 MB on all of the five products tested. Only Western is a little behind with less than 150 MB /s, the rate of an SATA 1.5 Gbits /s interface, while we are in SATA 3 Gbits /s with a theoretical maximum of 300 MB /s. Either way this max is far from being reached.


This graph represents the rate in KB /s depending on the advancement on the hard drive expressed in GB /s (X axis). The evolution since the E7K500 and its 100 GB platters is obvious and the Caviar SE16 dominates followed by the Samsung T166. You may have noticed that the rate at the beginning of the platter isnít standard which leads us to believe that there is a lower quality of media on the part of the drive tested. The Hitachi and Seagate share the back of the pack with drives from the last generation, the Hitachi being faster at the beginning and the Seagate finishing stronger.


Here the maximum, minimum and average are given of the same data.


We finish this section with access time measurements still carried out in h2bench. This was done first without noise control and then with it activated. Unfortunately, there hasnít been any improvement in this area and all hard drives are in the 1.2ms range. We point out the very significant impact of AAM for Western and Hitachi. On the other hand, with Samsung the loss is small, but this should be taken into account with the sound measurements provided on page 8. As usual, Seagate doesnít offer this option.


Page 4
Performances Ė Applicative index

Performances Ė Applicative index
We now move on to less synthetic tests, starting with a performance index made from results obtained for the ę XP Startup Ľ, ę Application Loading Ľ and ę General Usage Ľ profiles of PC Mark 2005. This involves an HDDís activity in starting Windows XP (90% reading, 10% writing), in launching various applications (83% reading, 17% writing) and normal use of a PC (60% reading, 40% writing). The weight of each one of these profiles in our index figure is 1/6, 1/3 and 1/2, respectively.


Here, we canít really see any improvement in performances since the E7K500 and the hard drives without AAM (audio acoustic management) are found at intervals of 10%. The Hitachis are in the lead, followed by Western, Seagate, and then Samsung. The activation of acoustic management doesnít change the ranking.


Page 5
Performances in file copying

Performances in file copying
We now move on to file copying. We looked at the rates in reading, writing as well as in copying with a combination of two large files totaling 4.4 GB, plus 2620 files of 2 GB, and finally 16,046 with a total of 733 MB. The source or the target in reading or writing on the hard drive was a RAID of two 74 GB Raptors capable of assuring a rate of 110 MB /s in order to not be limited on this side.


The Samsung T166 was the fastest followed by the Western, which however had better synthetic performances in sequential reading. The 7200.10 confirms that gross performances arenít the only factor, because in terms of reading and writing, itís in last place while the E7K500 is ahead despite its brute rate that is 9 MB /s less.

There is also an advantage for Samsung in copying files on the same hard drive, whether itís in close (in the same partition) or far (on a partition that starts in 50% of the hard drive) copying. Next are Western, then Hitachi and Seagate, depending on if the copying was close or not.


Activation of sound management doesnít change the ranking and we can see that the impact on performances is relatively insignificant even when itís a far copying and the reading head has to regularly go over half of the hard drive.


Page 6
IOMeter performances

IOMeter performances
IOMeter is used to simulate the load in a multi-user environment, in this case by using a server file load composed of 80% reading and 20% writing all 100% random on the hard drive. Here, the NCQ can be particularly useful because of the multiple numbers of simultaneous commands. In this test, we measured performances expressed in inputs /outputs per second (IO/s) with 1, 2, 4, 8, 16, 32, 64 and 128 simultaneous commands. Of course with a single command the NCQ has no effect.


One hard drive manages to partially do well here, the Western Digital. While it matches the ę good oleí Ľ E7K500 in a single command, its advantage increases with the number of simultaneous commands. This is proof of better NCQ management. In second place is Seagate, followed by the E7K500, the T7K500, and finally Samsungís T166.


With this type of use, the impact of sound management on performances can be very significant. With a single command, the effect on the Hitachis is 25%, versus 11% for Samsung and 29% for Western Digital. With 128 simultaneous commands, the impact is more limited, because this figure is down to 14%, 4% and 15%, respectively. Knowing that the Western is the one whose access is the most effected by AAM, contrary to Samsung, these results are more than logical. However, despite this reduction, Western is ahead as soon there are two simultaneous commands managed by the hard drive.


Page 7
Consumption and temperature

Consumption

Here we have the power consumption for each hard drive using an amperametric clip on the 5V and 12V lines of each model. Logically, itís the E7K500 and its five platters that is the highest while the others are quite close with small advantages for Samsung and Western.
Temperature

Next we measured the hard drive temperature after two hours of intensive use with IOMeter. These results are with the HDD out of its casing, without ventilation, and each slightly raised. Temperature is taken using an infrared thermometer in two areas; in middle on top of the drive and on the hottest part on the left side. Finally, we also give you the temperature as reported by the driveís internal probe.

The Hitachis are clearly those that heat up the most regardless of the generation. On the other side is the Western drive with the best figures even if its probe indicates 2įC more than the Samsung.


Page 8
Noise management

Noise management
For this measurement we place each hard drive in a Textorm Jelly Case in order to avoid any vibration from the desk. A sound probe is placed 12cm above and the PC used is of course fanless so that we can isolate the hard driveís sound level:


None of these models can really be called ę silentĽ and when combined with fanless solutions they will be slightly noticeable with their rotation alone. Only mono-platter drives will fit into this category, and we are still (of course) far from SSD with flash memory.

In rotation, the E7K500 and its five platters is slightly noisier, but it is especially in access that the difference can be heard. Without acoustic management, the Seagate 7200.10 is noisy, just like the Hitachi E7K500. The T7K500 as well as Western and Samsung models are more silent, though far from being discrete, with the T7K500 emitting the least amount of sharp sound.

The activation of acoustic management allows a noticeable reduction in noise coming from drive access. For Hitachi and Western, access doesnít really increase noise, which isnít the case for the Samsung, whose access time is also decreased with AAM. Moreover, while Samsung and Western have similar dBA readings, in practice the Western is more discrete in access. Finally, Seagate still doesnít offer AAM and 7200.10 owners will have to endure the noise.

To have an idea of the type of noise emitted by these drives, here are some recordings taken with a portable Roland Edirol R-09 recorder placed 12cm above. 15 seconds in rotation were measured followed by 15 seconds in access with acoustic management (except for Seagate) and finally 15 seconds without this option.

- Hitachi E7K500
- Hitachi T7K500
- Samsung T166 - HD501LJ
- Seagate Barracuda 7200.10
- Western Digital Caviar SE16 Ė WD5000AAKS



Page 9
Conclusion

Conclusion
There is something we didnít mention yet in this article, mainly, warranties and reliability. In terms of the first point, itís three years except for Seagate which is five years. This doesnít mean that these hard drives are the most reliable, but this is a good point that some will find important.

It is still as difficult to obtain trustworthy after sales figures, all the more so given that amongst these four manufacturers, Samsung doesn't allow direct returns and you have to go through the retailer. For this reason, a retailerís figures wonít be correct as they necessarily look bad for Samsungís products. Thus for a large French distributer, figures for Hitachi, Samsung, Seagate and Western are respectively 1.1, 2.3, 1.2 and 1.2% over one year.

And our opinion on the results on this test? In terms of performance, itís difficult to choose. Hitachi leads in PC Mark, Samsung is ahead in copying files and Western has the better results in IOMeter. The Western Caviar SE16, however, is always well situated and is therefore the most consistent of these models. In addition, there is its high performance NCQ, which allows it to have an advantage in multiple simultaneous access.

As for the ecological aspects, the Hitachi models seem to heat up more than the average hard drive, while the Western is the best in this domain. For noise, Seagate is, as usual, in last place, because its 7200.10 is very noisy in access and acoustic management still isnít available. Samsungís T166 is the most discreet in rotation, and access is comparable to the Western Caviar SE16 and Hitachi T7K500 (even if we prefer the latter because sounds arenít as sharp).

We do find it unfortunate that acoustic management isnít more effective for Samsung in exchange for slower access time like with the competition. And actually, here this is more audible than for the Western and Hitachi drives as their access is almost entirely masked by the sound of rotation.


In the end, if we were to recommend a single product, it would be the Western Caviar SE16 (WD5000AAKS) just for its consistent result in tests. Just behind are the Samsung T166 and Hitachi T7K500, which are distinguishable by their respective sound levels in rotation and access without AAM. The 7200.10 from Seagate has the sole advantage of its five year warranty which isnít too convincing compared to its noise level.


Copyright © 1997-2014 BeHardware. All rights reserved.