by Andrey Kuznetcov
12/14/2003 | 12:25 AM
The external hard disk drives will hardly surprise anyone today. Many well-known HDD manufacturers started expanding their product range lately by adding solutions like that. And there is every reason for this, actually.
After the USB 2.0 interface has completed its triumphal invasion into the users’ PCs, it appeared quite possible to take full advantage of the high HDD speeds even if they are implemented as external solutions. And if we also recall the FireWire interface, which is available in many today’s mainboards or can be acquired due to Audigy sound solutions or special add-on controller cards, it will be absolutely clear that the external hard disk drives are nothing exotic already. External HDDs no longer indicate how wealthy or “progressive” the user is, but can make users’ life a lot easier in many situations.
It is true that with a multi-gigabyte storage device, which can work efficiently with any modern computer, you can be absolutely sure that no unpleasant surprises occur. If you need to transfer a lot of data or create a backup of vitally important information, the external HDD solution will save you a lot of time and effort.
Besides classical hard disk drives intended to be installed inside a PC, Maxtor Company also got into external HDD manufacturing. Of course, we couldn’t disregard this fact that is why today we are happy to offer you a review of the fastest solution from Maxtor OneTouch family aka Personal Storage 5000DV.
The entire external hard disk drive family from Maxtor is called OneTouch. Where does this name actually come from? Well, if you take a closer look at the 5000DV HDD (Personal Storage 5000 Family), you will notice a large round button on the left of the front panel. With the help of this button you can create a backup for all files on your computer: just press it once. To be able to do it you will have to install special software enclosed with the drive. Besides that the front panel also carries two LEDs indicating the work mode status. At the bottom of the case there are four small legs, which raise the drive above the desk or any other surface it stands on and allow air circulation through the fan slits in the bottom of the case. At the back of the case there are two FireWire ports and a USB 2.0 port, as well as a power supply connector.
The 5000DV external HDD family includes three models with 120GB, 160GB (our today’s hero) and 200GB storage capacity. The first one features 2MB buffer, while the other two boast an 8MB buffer. DiamondMax Plus 9 used in this external solution supports ATA/133 interface. The spindle rotation speed is 7,200rpm and the average seek time the HDD provides equals 9.3ms. The peculiar thing about our today’s product is the multiple interfaces it supports. It supports simultaneously USB 2.0 and FireWire, which allow “plug-n-play” connection to the PC. As is known, the peak bandwidths for USB 2.0 and IEEE 1394 interfaces are 60MB/s and 50MB/s respectively. The actual maximum data transfer rate claimed by the manufacturer makes 34MB/s for USB 2.0 and 40MB/s for FireWire. The drive is not very small, it is 41mmx152mmx219mm, which means that you will not be able to put it into your pocket, however, this is a pretty standard size for a solution of the kind. The drive weighs 1.22kg. It supports a wide range of input voltages from 110V to 240Vand 47-63Hz current frequency at the temperature between 5 and 40oC.
Dantz Retrospect Express utility, whipped together with the Maxtor drive will help to make the backup process for your valuable data much simpler:
The external hard disk drive is shipped with a USB cable, 6-pin FireWire cable, external power supply unit with the corresponding cables, a special stand for the HDD to be set vertically, a CD disk with the software (Dantz Retrospect Express utility allows creating an automatic reserve copy of the data on the system drive by pressing OneTouch button), a user’s manual and a quick installation guide.
To test the actual performance of this external hard disk drive solution we used the following benchmarks:
Our testbed was configured as follows:
We tested the performance of our HDD with the help of USB 2.0 (ICH5) interface available on the mainboard, and FireWire interface provided by PCI adapters based on Agere FW323 and VIA VT6307 chips and working with the operation system drivers. For WinBench tests, the drives were formatted in FAT32 and NTFS as a single partition with a default cluster size. In some cases described below we used 32GB logical partitions also formatted in FAT32 and NTFS with a default cluster.
As usual we would like to start with the results our hard drive showed in DataBase pattern. In order to better understand what they mean, we built three graphs characterizing the HDD behavior in this test most vividly.
The first graph of the three shows the HDD performance under linear workload (that is when the queue=1). The performance difference is minimal as we can see, we could even say that there is almost no difference at all. Both interfaces prove equally fast here. I wouldn’t say that any of the IEEE 1394 controllers is preferable here, too. The higher gets the write requests share, the faster works the HDD, which indicates that it uses very efficient lazy write algorithms.
The second graph you can see, stands for queue depth equal to 16 requests. You can see from the pic that FireWire interface looks more attractive then USB 2.0 up to 90% writes share, when all the three graphs merge together. Both IEEE 1394 controllers perform almost equally fast. USB 2.0 controller doesn’t support concurrent requests, so that the performance of the drive working with USB 2.0 interface didn’t change as the queue depth grew higher. At the same time, IEEE 1394 controllers do have this feature. As a result, the HDD runs considerably faster at first, but as the writes share increases, the advantage gets less and less evident, until all three graphs merge together. Judging by this fact I dare suppose that FireWire controllers know to optimize the read requests order.
As the queue depth increases up to 256, we can see how big the advantage of the FireWire interface over USB 2.0 is. It is especially noticeable when the writes share is still pretty low. The performance difference between the two IEEE 1394 controllers is again negligible. As the writes share grows up, the performance gap between the two interfaces reduces.
During the SequentialRead and Write test we check how well the hard disk drive can cope with ordered requests featuring linearly growing address. The requests queue depth in this case remains equal to 4, and once per minute the size of the requested data block changes.
The SequentialRead graph is an excellent piece of evidence confirming the specifications claimed by the manufacturer. You can clearly see from this graph that the curve standing for USB 2.0 interface crawls up until the data block size reaches 64KB. Then it turns into a horizontal line indicating that the HDD performance is limited by the interface bandwidth. As we have already said the manufacturer claimed it to be 34MB/s for USB 2.0 interface.
The same thing is true for FireWire interface. The HDD ran fastest of all when the data block size is 1MB, having reached the claimed maximum of 40MB/s. Again, both IEEE 1394 controllers run equally fast.
During SequentialWrite tests the HDD runs slower, than during erading, as we have actually expected. Again USB 2.0 looks less attractive. The write speed when the HDD work via this interface reaches the promised maximum, which is slightly above 26MB/s, with the 64KB data block (you can see a corresponding graph running horizontally at a certain point). With FireWire interface, the hard disk drive performs more impressively. Here we can already notice the performance difference between configurations with different controllers involved. You can already see it in case of 4KB data blocks and up. The HDD shows better write speed when it is used with Agere FW323 controller, which reaches almost 2MB/s.
Let’s see how fast our external hard drive will be when working with server patterns:
When the hard disk drive works under FileServer workload, FireWire interface becomes more attractive than USB 2.0. In case of 1 and 4 outgoing requests the advantage is stable, though not that huge. However, as the number of outgoing requests grows up, the performance gap also increases and by the time we reach 256 requests, the HDD performance more than doubles. The two IEEE 1394 controllers run equally efficient in this case.
When this hard disk drive is used as a part of a Web-server, the picture is generally the same. Under linear workload the interface type doesn’t matter for the drive, however, as the number of processed requests reaches 4 already, FireWire controllers speed up the processing. Just like in the previous pattern, we cannot name any of the two IEEE 1394 controllers the winner.
The work of our hard disk drive under typical workstation load is limited to the maximum of 32 requests. The results were taken for the entire HDD as well as for the first 23GB of the storage space.
The final graph shows that with USB 2.0 interface the performance of the drive starts growing pretty rapidly in the beginning and then reaches the defined maximum getting more or less leveled out.
FireWire interface looks much more preferable in this case. Despite the increase in the queue depth in the beginning, the HDD performs stably fast, which you can see from the horizontal part of the graph. As soon as the queue reaches 8 requests, the drive speeds up quite notably. The two IEEE 1394 controllers perform equally fast, which you can see from the nearly coinciding graphs.
The case when only one 32GB logical partition of the entire HDD is used looks very similar to what we have just seen on the pervious graph. The difference primarily lies in a much higher level of performance, which is caused by the 32GB address space limitation of the work zone. The second peculiarity is a slightly different shape of the graph for USB 2.0 interface, which is nearly horizontal.
When we connect our external HDD via FireWire interface, the requests processing speed turns out higher again. And again the HDD performance boosts when we reach 8 requests queue depth. Both IEEE 1394 controllers are equally efficient here.
We tested our hard disk drive with the help of File Copy utility following the standard testing methodology. As usual, we used five standard file sets. Besides the main logical partition, 32GB big, there was also another one, of the same storage capacity, which was used to save the files copied to it from the first main partition.
During file creation FAT32 file system demonstrates much higher speeds with USB 2.0 interface. FireWire loses to USB 2.0 being 2-3 times slower in all tests. This significant lag behind makes me suppose that the internal controller of the drive responsible for FireWire-to-ATA bridge is not efficient enough for operations of the kind. The two IEEE 1394 controllers perform equally fast. We could suspect bad work of the FireWire controller driver, however, as you remember we used the same default FireWire driver for the WD hard disk drive tests and didn’t encounter any problems with data writing. Moreover, in our today’s test session the results of the synthetic SequentialWrite pattern indicate that Maxtor drive is faster with FireWire controllers. All in all, it’s a mystery.
Even worse results are obtained with FireWire interface under NTFS file system. The HDD creates files at about the same speed as in the previous case when it works with USB 2.0 interface, and in case of FireWire the performance is lower than under FAT32. As a result, the performance difference between two interfaces starts notching 4 times for those cases when we have a bunch of smaller files. The performance differences between two IEEE 1394 controllers are again insignificant.
File reading under FAT32 file system works faster with FireWire interface. In this case the HDD manages to speed up quite tangibly than in case of USB 2.0 interface. Again both IEEE 1394 controllers perform close to one another.
In NTFS file system file reading is again faster with FireWire interface than with USB 2.0. both controller for IEEE 1394 are equally efficient as usual.
File copy within one partition formatted under FAT32 occurs faster when the external drive is connected via the FireWire interface. The performance with USB 2.0 interface appears much lower.
The NTFS file system during file copy demonstrates the indisputable efficiency of FireWire interface when working with a bunch of relatively large files. In these situations the storage device works as fast as with the USB 2.0 interface. And in case of ISO file set it even shows its best. As the number of files in the test sets increases (Windows and Programs patterns), the performance of our hard drive with FireWire interface turns out notably lower. The two IEEE 1394 controllers are equally fast.
File copy from one logical drive to another under FAT32 indicates that FireWire works better in this case: it wins in four cases out of five. It loses only in case of the largest files processing. Both IEEE 1394 controllers run neck and neck, so that we cannot give our preferences to any of them.
In NTFS file system the file copy speed from one logical drive to another is about the same as in FAT32. When working with relatively large files the copy speed grows up a little bit, and when we get to a bunch of smaller files, the copy speed drops. USB 2.0 interface appears more attractive in all cases. Its advantage over FireWire is especially evident in those cases when we process a bunch of smaller files. The two IEEE 1394 controllers retain certain parity.
The last test we are going to run on our external drive is the classical WinBench99. The tests were run for a HDD formatted as a single logical drive as well as a 32GB logical drive.
First please check the linear read speed graphs:
Well, let’s start with the read speed from the HDD surface for different interfaces. All the three graphs we got, illustrate perfectly well the claimed peak bandwidths of the interfaces. Long horizontal lines characterize the peak bandwidths of each interface. With its maximum performance of about 60MB/s in the beginning of the drive, the HDD can’t push its linear speed beyond the limit set by the interface used. As you can see from the graphs, the top limit for FireWire interface is about 6MB/s higher than that for USB 2.0 and corresponds exactly to the values claimed by the manufacturer. The two IEEE 1394 controllers, Agere FW323 and VIA VT6307, perform about the same.
The tests of the hard disk drive in FAT32 file system demonstrated indisputable advantage of the USB 2.0 interface: the two key scores, namely Business Disk and High-End Disk WinMark appear the maximum in this case. FireWire interface looks less attractive here, and in case of VIA VT6307 controller the results are the worst: this solution yields quite tangibly to its counterpart on Agere FX323 chip in these two major tests.
The work of our hard disk drive under NTFS file system is evidently slower on both interfaces. The advantage of USB 2.0 over FireWire is even more obvious in this case. Both Disk WinMark scores are about twice as high with USB 2.0 than with FireWire. Agere FW323 controller is again slightly faster in both these parameters than the solution from VIA.
When we work only with a 32GB logical drive, the results shown by our HDD are definitely higher than in case the entire storage capacity of the drive is involved. In FAT32, USB 2.0 interface provides much better scores in Business Disk and High-End Disk WinMark tests. The two IEEE 1394 controllers retain the results parity, so that we cannot name the winner at all.
The use of 23GB logical drive for the tests in NTFS file system again indicated USB 2.0 interface as the most preferable in this case. Among the IEEE 1394 controllers, the best one is again Agere FW323: with this controller the results in Business Disk and High-End Disk WinMark are notably higher.
Well, let’s sum up the results now. I would like to start with the impression left by the HDD first. I have to admit that the drive’s functioning didn’t cause us any troubles, so that I would call this experience highly positive (here I have to disregard the unpleasant surprise with the write speed in FC-Test with FireWire interface). This external HDD model will be a good choice for those of you who need a large portable storage device for storing a lot of data, which boasts pretty high speed characteristics. The availability of two efficient interfaces, such as USB 2.0 and FireWire makes the application field for this solution much wider and provides it with high level of flexibility, which is essential nowadays for maximum performance. The accompanying software bundle allows simplifying the valuable data copy, so that it is done in the most optimized and effective way. The read speed scores obtained throughout this test session prove the manufacturer’s claims regarding the practical bandwidth of the USB 2.0 and FireWire interfaces.
Now I would like to say a few words about the highs of the supported interfaces. Despite the theoretical advantage of the USB 2.0 interface over FireWire (60MB against 50MB), the latter appears more preferable in real applications. It is true, IEEE 1394 allows transferring the data at 40MB/s while High-Speed USB supports data transfer at 34MB/s only. However, far not all the tests indicated higher efficiency of FireWire interface. The reasons for that probably have to do with the not quite adequate work of the FireWire-to-ATA bridge (or its incorrect work with the FireWire controllers we used).
It is very likely that the HDD performed the same way with both IEEE 1394 controllers. They looked like twin-brothers in most benchmarks, that is why it was really hard to single out the winner. Anyway, I would consider Agere FW323 to be a better choice. Nevertheless, let’s not make any final conclusions and do a little bit more research on that in our ongoing reviews, because we haven’t answered all the questions today, for sure. Of course, it is totally up to the user to decide which interface to choose, and will sometimes be determined by the availability or absence of this or that interface on the user’s PC.