6 HDD SATA from 500 to 750 GB - BeHardware
>> Storage >> Hard Disks

Written by Marc Prieur

Published on June 16, 2006

URL: http://www.behardware.com/art/lire/628/


Page 1

Introduction



Do you need huge storage capacities? If you are ready to put your trust in a single drive to store all your vacation movies, all manufacturers except Samsung offer 500 GB hard drives. Seagate even sells a 750 GB. What is the performance of these monsters?


3, 4 or 5 platters !
The 6 drives tested here conform to SATA 3.0 Gbits /s standards and support NCQ. We described these two technologies in our 250 GB drive survey a month ago. They all also have 16 MB of cache.


From an architectural point of view, the main difference between these products is the number of embedded platters to reach 500 GB. The Hitachi 7K500 has five 100 GB platters compared to four of 125 GB for the Seagate Barracuda 7200.9, Maxtor DiamondMax 11 and Western Caviar SE16 WD5000KS.


Thanks to perpendicular recording, Seagate reaches this capacity with three 167 GB platters for the Barracuda 7200.10. This is the reason why only 4 platters are necessary for Seagate to reach 750 GB for the same product line. They are the first to introduce to 3"1/2 SATA drives discs using this type of recording and it noticeably increases surface density. Other manufacturers will soon release similar products. If the Hitachi T7K500 (which isn´t available yet) doesn’t use it, Hitachi has already released a laptop hard drive product line using this technology.



Page 2
The discs, the tests

Pictures of the hard drives
Here are the hard drives used for this test:

- Hitachi 7K500 500 GB
- Maxtor DiamondMax 11 500 GB
- Seagate Barracuda 7200.9 500 GB
- Seagate Barracuda 7200.10 500 GB
- Seagate Barracuda 7200.10 750 Go
- Western Digital Caviar SE16 WD5000KS 500 GB

Hitachi 7K500Maxtor DiamondMax 11
Seagate Barracuda 7200.9 Seagate Barracuda 7200.10 500 Go
Seagate Barracuda 7200.10 750 Go Western Caviar SE 16 WD5000KS
There isn´t much to say about the front picture except that one product is black. Manufacturing dates of the hard drives tested vary from December 2005 to March 2006 for those explicitly indicated.

Hitachi 7K500Maxtor DiamondMax 11
Seagate Barracuda 7200.9Seagate Barracuda 7200.10 500 Go
Seagate Barracuda 7200.10 750 GoWestern Caviar SE 16 WD5000KS

There is no real surprise for controller cards, as chips are the same as those used for lower capacity drives (except for Seagate). The two 7200.10 include one ST controller chip instead of the usual Agere. The two chips however have the same pin configuration and the PCB of the 7200.9 500 GB and 7200.10 750 GB are similar: we could even think that this is actually the same chip. Maxtor uses Agere´s controller, Hitachi uses Infineon´s and Marvell the Western. Only the Hitachi and Western drives have a standard Molex connector in addition of course to the standard Serial ATA power supply.
Test protocol
For this test, we have used the same protocol as for the previous 250 GB test, except that we only used an Intel platform. First we looked at the drives “synthetic” performances; cache and sequential transfer rates and average seek time. Then there were more applicative tests, which include an applicative performance index based on PCMark05, a simulation of a file server type of load via IOMeter. And also a writing, reading and close copy (in the same partition) and far copy (starting at 50% of the drive) of a group of files.

These files include 2 big files for a total of 4.4 GB, plus 2620 files for a total of 2 GB, and finally 16,046 files for a total of 733 MB. The source or target for reading or writing on the disc are two Raptor 74 GB in RAID, capable of 110 MB/s transfer rates without restrictions in this area. Of course, this is of little interest, because if the sequential transfer rate gives an idea of performances during the copying of big files, things are different with small ones.

All measurements were taken with drive sound management deactivated and then at times with this option activated (except for Seagate, because the function isn’t supported). We also measured the performances of the chipset.

With the i975X, we measured with and without AHCI mode activated. With standard parameters, Intel’s controller is configured in standard IDE and the advantages of SATA such as NCQ or hot plug aren’t supported. It has to be configured in AHCI (Advanced Host Controller Interface) to benefit from these technologies.


Page 3
Synthetic performances

Synthetic performances
First of, we start with cache data transfer rate measurements made with h2benchw and the « Core Test ». All tests were made with hard drives configured in SATA 3 Gbits /s. This isn´t the initial configuration of the Seagate 7200.10 (a jumper had to be removed) and the 7K500 (to be activated via Feature Tool). The Seagate drives have the fastest cache, followed by Hitachi and Maxtor. The Western’s performances are low since it only exploits half of the available bandwidth.


The sequential transfer rates of 500 GB drives, represented here from the first to the last GB, is surprising, because it isn´t in complete correlation with the density announced by manufacturers. The 7200.10 is the fastest followed by Western. Also, if the Hitachi drive and the 5 platters of 100 GB is logically one the slowest, it’s equivalent to the 7200.9 and its four platters.


Now here are the maximum, average and minimum sequential transfer rates for all hard drives. The 7200.10 is the fastest with 60 to 61.8 MB /s on average and 74.9 to 76.3 MB /s at top speed: the 750 GB is logically a little faster than the 500 GB since it uses platters of a higher density (187.5 GB as compared to 167 GB). Western is just behind followed by Maxtor. With 47 MB/s on average, the previous generation of hard drives such as the 7K500 and 7200.9 are clearly slower.


For seek time, however, the 7K500 is much better and finishes in first position. Maxtor is last and the Western and Seagate are approximately equivalent.


Page 4
Perfs - applicative index

Performances – applicative index
We start with the less synthetic tests and a performance index based on figures obtained with « XP Startup », « Application Loading » and « General Usage » profiles of PC Mark 2005.


The first thing to notice is that the Maxtor and Hitachi benefit the most from activation of the AHCI mode. A second point is the low performances of the Maxtor when sound management is activated! There is on impact on performances for Hitachi, which isn´t the case for Western. Overall, the Hitachi drive wins followed by Western.


Page 5
Perfs - file copying

Performances – file copy
The next test is file copying. We measured reading, writing, and also the copying of the following files on hard drives: 2 big files for a total of 4.4 GB, plus 2620 files which total 2 GB, and finally 16046 files which weighs 733 MB. The source or target for reading or writing on the disc are two Raptor 74 GB in RAID which are capable of ensuring a transfer rate of 110 MB/s (there is no performance restrictions on this side).

This type of information is uninteresting, of course, because if the sequential transfer rate gives an idea of performances during the copying of big files, things will be much different with small ones. We followed two procedures to copy the files: writing in the same partition in the beginning of the disc and from this partition to a second one that begins at the middle of the disc.


The sequential transfer rate is very helpful here and we end up in the end with the Seagate 7200.10 Barracuda generation in the lead. Despite a higher sequential transfer rate, the Western drive ends up behind the Maxtor. The impact of automatic acoustic management (AAM) is noticeable without being too significant whatever the disc.


For writing, the data transfer rate falls and sometimes using AHCI doesn´t help. This is the case for the 7200.10 or even the Maxtor without AAM. However, gains are very significant for the WD5000KS. Without AHCI, the 7200.10 are in the lead, but with this option activated, Western is the fastest.


For file copying, the impact of the ACHI is once again variable depending on the disc. For Hitachi and Maxtor, for example, the AAM needs to be activated to measure a slight performance gain. However, Seagate and most of all Western always benefit from its activation. The 7200.10 is the fastest.


Page 6
Perfs - IOmeter

Performances – IOmeter
IOMeter simulates a load in a multi-user environment. It’s a load of file server type comprised of 80% reading and 20% writing, which is 100% randomly accessed on the disc. In this case, the NCQ can be particularly useful. We tested the IOMeter with a number of concurrent commands from 1 to 128 and of course for a single command, the NCQ doesn’t bring any improvements.


It’s true, the impact of NCQ isn´t the same for all hard drives. For example, the gain isn´t really significant for the 7K500. This is something that we already noticed with the T7K250. For Seagate, there is a slight performance gain and it’s with the Maxtor that it seriously starts to be appreciable. Indeed, if it was in last position without NCQ, it finished second when it’s activated. The WD5000KS however, already in the lead without NCQ, is untouchable with it because of the activation of Intel´s AHCI controller.

If you carefully look at the two graphs you will notice that with 1 and 2 concurrent accesses the 7200.10 750 GB performance is slightly inferior with NCQ. We ran tests several times and always came up with the same results without any apparent explanation. This is all the more incomprehensible that with the 7200.10 500 GB this doesn’t happen.


Now, here are Hitachi´s performances for the four different configurations available for this test. The AAM has a negative impact on performances, but we noticed that as the number of concurrent commands increases, the impact is less. Strangely without AAM and in AHCI, the drive is faster with 32 accesses than with 64 or 128.



If it already showed very low performances in our first application test, the Maxtor drive once again shows extremely low performances with AAM. This time the only reason is that Intel´s controller isn´t in AHCI. The performance gain isn´t as fast as the number of commands increases, and beyond 16 it even tends to decrease. Except for this point, results are satisfactory.


The activation of AAM with the Caviar SE16 seriously reduces performances with IOmeter especially if the NCQ isn´t activated. In the same case, the Hitachi is even faster. The western drive is once again in the lead with NCQ+AAM compared to the 7K500 and the same configuration beyond 32 concurrent accesses because of a much more efficient NCQ.


Page 7
Noise and thermal dissipation

Noise and thermal dissipation
We saw in previous tests that the impact on performances due to the activation of AAM was the lowest for Hitachi and the opposite for Maxtor. It’s so great for the latter that we even think that there is a problem with the firmware. What about seek time?


Logically, it’s with Hitachi that the AAM increases seek time less, followed by Maxtor and Western. At 21.6 and 21.7 ms for the two drives, this starts to be very high.

A hard drive test with the AAM function would be incomplete without a measurement of noise levels here taken at a distance of approximately 2 inches.


Just for spinning there are very significant differences. So, if the Western and Seagate are relatively discreet, they aren´t however silent. This is less the case of the Hitachi and Maxtor, however, their noise level increases less during seek while they are still noisier than the Western. For once, the Seagate drives are the noisiest in intensive seek and AAM can do nothing as this option isn´t available.

It is relatively difficult to know which drive dissipates more heat. The internal disc probe isn´t necessarily located at the same place and the hard drive doesn´t heat uniformly. This last point is also problematic if we want to place a thermometer at a specific location on the hard drive.

However, many of you wanted to have more information in this area. In order to obtain data that gives us an overall idea, we conducted an original new test. Each disc has been placed in a 8"x 6"x5.5" closed box equipped with a thermometer to measure air temperature with a small hole for the power supply and SATA input.

After one hour of intensive work with IOmeter, we measured the air temperature increase inside the box compared to ambient air. This is how we found out which drives dissipate the most heat:


The difference is relatively low at 1.9°C, which represents a 26% gap. The discs that increase air temperature the least were the Western and Seagate 7200.9 500 GB. The opposite result is provided by the Seagate 750 GB.

For comparison, here are the temperatures measured by the drives´ internal probes and obtained via SMART during this test in very extreme conditions:



Page 8
Conclusion

Conclusion
With 500 to 750 GB of storage space, or 465 to 698 GB with 1 KB = 1024 bytes, these drives will be used in several different ways. For some, this will be a single drive used like a system and storage space. For others, it will only be devoted to file storage, which will generally be very big and accessed one by one. It will also be possible to find them, for example, in a NAS and accessed by many users simultaneously.

For single user storage, the 10th generation of Seagate drives is a good choice. They have low spinning noise level and have the most significant transfer rates because of the record density. Seagate is the only one to sell a 750 GB drive.


For system drive use or a file server, our preference goes to the Hitachi 7K500 and the Western WD5000KS. If the Hitachi has superior applicative performances and more important data transfer rates for file copying without AHCI, the Western is faster with AHCI for file copying and provides much higher performances for a multi user environment.

Finally, the Maxtor DiamondMax 11 is cheaper. It is usually found at about 250€ instead of 280 to 300€ for the other 500 GB drives. Even if it isn´t the best, it is the most affordable. This is however its only advantage, all the more that it provides from time to time abnormally low performances when AAM is activated.


Copyright © 1997-2014 BeHardware. All rights reserved.