Hard disk drive is one of the important components of every system. However, you are very often at a loss, which one to choose? Today we decided to offer you a detailed comparative coverage of the peculiarities and performance of 7 different hard disk drives from the leading manufacturers. All of them feature 20GB platters and their capacity makes also 20GB. This is not just a crazy idea of ours, but an attempt to make all the testing participants perform in similar conditions. As we have already stated in our previous reviews, the HDD fastness in certain tests is greatly dependent on the cluster size (and the latter depends on the HDD storage capacity when formatted in FAT32).

So, what are our today's heroes? We will take a closer look at the models from such well-known manufacturers as Quantum, Seagate, Western Digital and Fujitsu. Unfortunately, we didn't include the hard disk drives from IBM here (we are especially anxious to see the models from 60GXP series). However, we couldn't postpone this review any longer…

All the hard disk drives reviewed are already available in retail that's why we hope our recommendations might be very helpful to you :)

Testing Participants

For our review we selected hard disk drives featuring different rotation speeds that's why it would be incorrect to compare all of them with each other. In order to make the story simpler and not to overload you with categories and subcategories, we placed faster disk drives together with the slower ones on the diagrams and in the tables. The first group of devices featuring 7200rpm rotation speed includes: Quantum FB Plus AS, Seagate Barracuda ATA III and WD200BB. To the second group with 5400rpm (and lower) rotation speed belong: Seagate U5, WD200EB, Fujitsu MPG3204AT and Quantum FB lct20. Although the last HDD represents a particular type of hard disk drives, we decided to include it in our comparison as well. It should be of interest to you :)

Well, to begin with we would like to introduce the already familiar to you Seagate drives. We got two newest models at our disposal:

Seagate U5

This HDD is a budget solution from Seagate. Despite its relatively modest features, it proved pretty fast in benchmarks and operated almost noiselessly. See more in our Seagate U Series 5 HDD Review.

Seagate Barracuda ATA III

This hard disk drive belongs to the latest generation of IDE-drives. It features the today's maximum data density per platter, 20GB, and the rotation speed is 7200rpm. Altogether these two features offer us the data transfer rate of 500Mbit/sec, which is quite cool. The HDD is equipped with 80MHz DSP-processor and a 2MB cache buffer. Moreover, it can boast 8.9ms average seek time (reading), which is really small for HDDs with the data density that high.

Seagate Barracuda ATA III is available in two versions: with FDB (Fluid Dynamic Bearing) and with regular motors. So, the users can either buy a quieter HDD or a cheaper one with higher noise rate.

Of course, this hard disk drive is also provided with the Seagate's brand 3D Defense technology (read more about it here). You can find a hard disk with any of the following storage capacities: 10, 15, 20, 30 and 40GB.

Quantum Fireball Plus AS

As you can see, the HDD has hardly got a new looks. However, inside the package, there is a totally new fresh ATA/100 electronics, a 20GB platter rotating with 7200rpm and … (the most promising thing!) the average seek time is announced still to be equal to 8.5ms! Although this HDD can boast good technical characteristics telling positively on the performance, Quantum claims that it will be pretty quiet due to Hydrodynamic Bearing (HDB), which will be used only for single- and dual-platter models. Besides, this HDD is also equipped with a special Quantum's brand Quiet Drive Technology (QDT), aimed at reducing the noise level.

Moreover, Shock Protection System II technology is responsible for data protection. Namely, it minimizes problems associated with hard disk drive mishandling that typically occur before the drive is installed in a system, as well as those associated with shock instances during drive operation. SPS II prevents these events from having a detrimental effect on the drive's long-term reliability and data integrity. There is also a special Data Protection System, which tests the health of Quantum desktop HDDs. It helps prevent avoidable data loss. By determining whether the drive is the source of a system-level failure, DPS spares users from unnecessarily removing and replacing a perfectly healthy drive.

Fireball Plus AS models are available with the following capacities: 10, 20, 30, 40 and 60GB.

Quantum Fireball lct20

To tell the truth, the features of this HDD look not very impressive taking into account that we are living in the third millennium already. Just imagine: 4500rpm rotation speed, 128KB cache buffer… However, the 20GB platter and the cost of this device make it occupy a certain niche in the today's storage market. You can express your opinion about Quantum's decision to deal with the Low-End market, but you cannot condemn it…

Bearing in mind that Quantum has recently merged with Maxtor, we can't predict how long the lifetime of this disk drive will be. However, while it is available, the results it shows can be of interest to you.

By the way, this HDD is also provided with all the modern Quantum technologies mentioned above: QDT, SPS II and DPS. The product range includes 4 models: 10, 20, 30 and 40GB ones.

Western Digital Protege (WD200EB)

This hard disk drive features 5400rpm rotation speed and supports ATA/100 storage interface. Its cache buffer is 2MB big (compared to 512KB by Seagate U5 and 128KB by Quantum Fireball lct20).

In order to protect the electronic elements against accidental damage, they were removed from the outer side of the PCB inside the HDD. We were pretty surprised to find this out and worried about the drive's stability, because the overheating problem still exists and may be quite insidious.

A significant advantage of the new case of the WD Protege is the absence of the notorious pressurizing joint strip on the side of the drive.

By the way, if you didn't know the whole story about this strip, here you are. Numerous users and OEMs cursed the older WD hard disk drive model, because the strip it featured got torn when the HDD was installed into the 3.5" bay of the PC. As a result, the section with the platters appeared depressurized and dust could easily get inside the most vulnerable parts of the drive. The most interesting thing about it is the fact that WD refused to replace the out-of-order HDD with a new one or to repair it according to the warranty. In other words, WD was simply laying antipersonnel mines for their customers and when they stepped on these mines, Western Digital said that these were their problems… An excellent idea, don't you think so? :-)

As for the models available, there are only two of them: a 10GB and a 20GB one.

Western Digital Caviar (WD200BB)

WD Caviar 7200rpm drives offer Ultra ATA/100 and Ultra ATA/66 interfaces, 2MB cache buffer, low seek times (8.9ms) and up to 100MB/s data transfer rates. To protect your data there is a special Data Lifeguard tool, a combination of hardware features and software utilities. Data Lifeguard includes embedded features that guard your data and software tools that make installation, configuration and diagnostics simple and worry-free, as WD guys claim.

There are three capacities in the model row: 20, 30 and 40GB.

All in all, the specs look just impeccable. However, on reading this some of you may not agree with us, since WD hard disk drives have usually been famous for their stability issues. To tell the truth, it looks as if the situation is improving little by little nowadays. At least, the feedback we get from our readers shows that there are really few complaints about it now. New case design and few platters (heads) used in the HDD construction told on the overall stability of the device.

Fujitsu MPG3204AT

Frankly speaking, Fujitsu hard disk drives didn't prove the fastest in our previous tests. When we prepared for this huge testing we really hoped that Fujitsu would try to change this tendency with the new model introduced…

However, as soon as we saw the specs of this piece, we understood that this disk drive was one of the most mysterious ones. For instance, the disk may be built with Fluid Dynamic Bearings (FDB) or without them, because according to the specs it may feature "Optional Fluid Dynamic Bearings". Though there is no mention about the differences between the models with and without them.

The second surprise was the cache buffer size: 512KB, though here you may read: "Option: 2MB".

We got really puzzled by all this stuff. Does this mean that Fujitsu is keeping their fully-fledged models with all those cool features for the Japanese market?

Having taken a look at each of the today's heroes, we would like to offer you a comparative table, which should help you see the differences between the models considered:

Testbed and Testing Methods

Our test system was configured as follows:

• Intel Coppermine 600MHz CPU;
• ASUS CUBX-E bios 1007A;
• 2 x 128MB PC133SDRAM by Hyundai;
• Matrox Millennium 4MB graphics card;
• Windows 98/Windows 2000 Pro.

Now we would like to say a few words about our testing methodology.

In order to check the HDDs performance in different UDMA modes, we had to use the following controllers:

• UDMA33: the controller integrated into i440BX chipset;
• UDMA66: Promise Ultra66 controller;
• UDMA100: Promise Ultra100 controller.

For the Promise controllers we used the drivers ver. 1.60. In fact, these drivers were initially intended only for Promise Ultra 100, however, after a slight modification of the inf-file, they could be also run for Promise Ultra 66.

The disk drives were connected as Master-units to a separate IDE-channel. We used FAT32 and NTFS file systems to format each of them as one logical drive of the maximum size with the default cluster. DMA support in Windows was enabled. All the tests were run 4 times and then the average results were taken for the diagrams. The HDDs didn't rest for cooling down between the tests.

For our investigation we selected the following benchmarks:

• Windows 98:
  • WinBench 99 1.2
  • Adaptec Threadmark 2.0
• Windows 2000
  • WinBench 99 1.2
  • HDTach 2.61
  • IOMeter 1999.10.20

All the hard disk drives were supplied with the UDMA100-protocol enabled as default. In order to get more precise results and to avoid conflicting situations we used special utilities to switch the modes to UDMA33 and UDMA66 correspondingly. For your convenience we can offer you these utilities for download here. Some of them, however, can't work properly with the hard disk drive connected to an external UDMA controller that's why we would advise you to connect your HDD to the integrated controller before switching the UDMA mode with the help of these utilities. The main thing you should remember about is to make sure that you returned everything to the initial position :-)

So, you can download these utilities here:

• Quantum FB Plus AS
• Quantum lct20
• Seagate
• Western Digital
• Fujitsu

The newest and the most complicated test, which we applied to our review participants, was Intel IOmeter. You can check this article at Storage Review for more details on this test and the major testing principles.

So, now we are ready to start!

Performance

Access Time

It's not a secret that computer parts manufacturers are usually making the specs look somewhat more attractive than the thing really is, i.e. they mention those parameter values, which do not give the users any real idea of the hardware performance, but are impressively low. For instance, they do mention the average seek time in the HDD specs, instead of the more informative average access time:

Seek time is an average of how long it takes a drive to move the read/write heads to a particular track on the disk. It includes controller overhead but does not include drive latency.

Of course, this parameter is of interest to hardware engineers, but not to a user, because it doesn't illustrate the performance for him. The hard disk drive doesn't seek the track just for the sake of the process: it's goal is to provide the requested data to the user as soon as possible. The request consists not only of the track address. It also includes the sector address and the real time it will take the HDD to complete the task, which is a combination of time period required for the following operations: moving the heads to the required track + rotational latency + sector reading + data transferring. For the modern HDDs the last two items are relatively small and can be neglected. Rotational latency is taken to make half the time of the full disk rotation. When putting the track seek time onto the specs list, the manufacturers simply equate the HDDs with 5400rpm rotation speed with faster models rotating at 7200rpm. We will explain: provided the average seek time is the same in both cases, the drives with 5400rpm speed will have longer average access time due to higher rotational latency.

We've got two benchmarks in our set, which are capable of measuring the average seek time. This parameter is one of the key ones determining the HDD fastness that's why estimating its values for our testing participants "independently" really makes sense. We ran ZD Disk Inspection test and HDTach 2.61 tests in Win2000. Here are the obtained results:

As you can see, the values are very close to one another, however, they are somewhat different from what the manufacturers claimed. The most "unfair" model appears Seagate U Series 5, while by all the other hard disk drives the average access values are very close, nearly coinciding. Maybe there was no need misleading us then? It has already become true for CPUs that the customers buy megahertz, unfortunately...

Here is what we get from our testing participants in terms of average access time:

Quantum Fireball Plus AS is ahead of all in both tests, then a bit behind the leader comes WD200BB and then - Seagate Barracuda ATA III.

Since average access time depends not only on the fatness of the heads block but also on the platter rotation speed, it appears quite logical that those HDDs featuring 5400rpm spindle rotation speed fall behind their faster brothers with 7200rpm quite significantly. The only exception is Fujitsu hard disk drive, which fell just a little bit behind Seagate Barracuda ATA III.

Data Density

Data density is another factor telling on the HDD performance. You should keep in mind that not only the maximum and minimum values influence the performance, but also the size of a sector with even data density. For instance, if we take our favourite IBM DTLA HDD, we will see that the first sector features relatively high data density (the maximum media transfer rate makes 444Mbit/sec). Just compare it to the media transfer rate of HDDs with 20GB platters. In our today's competition, the hard disk drive with the highest density is Seagate Barracuda ATA III: 500Mbit/sec. Since the tests are run for the "fastest" sector in the very beginning of the hard disk drive, the results appear very high.

In order to measure the maximum and the minimum data density we resorted to Disk Inspection test from ZD WinBench package (the table below contains the results obtained for UDMA-100 mode):

When connected to the UDMA33 interface, some hard disk drives (almost all of them, in fact) behave quite "nervously" and the graphs turn out really interesting in this case. However, we were mostly interested in neat and clean graphs obtained with the higher protocol enabled. Only in this case the HDD limits its own performance.

According to the results we got we made up the following diagrams:

The indisputable leader here in terms of liner read speed is Seagate Barracuda ATA III. And the loser appeared the one with the lowest spindle rotation speed, of course, - Quantum lct20. What strikes as really unexpected here it's the performance of WD200EB: the graphs obtained for it were very similar no matter what UDMA mode it was and the maximum liner read speed turned out pretty low against the background of faster competitors.

Here the first prize belongs to Seagate Barracuda ATA III, and Quantum AS rolled down to the fourth place from the sacred second, having let forward WD200EB and… Seagate U5! Thanks to platters with higher data density Seagate U5 with 5400rpm rotation speed managed to leave behind an undoubtedly more powerful competitor, Quantum Fireball Plus AS.

WinBench99 for Windows 98

We have already used this benchmark to test the HDDs performance several times. That's why we won't describe its methods and algorithms. Let's pass over to the obtained results right away. We tested each hard disk drive four times for each UDMA-interface (you can imagine what huge job it was :-) Here is what we got:

Quantum Fireball Plus AS and Seagate Barracuda ATA III were running very close and wrested the victory from one another. Although Seagate managed to get more points in this competition. As for WD200BB, it proved really exciting here. It will never set the Thames on fire, however when working with Visual C++ files, it managed to leave all the rivals far behind. If you are a software developer, then think of getting yourself a WD200BB hard disk drive! :-)

Quantum Fireball Plus AS took a small revenge, so that the position of the slowest HDD of all appeared in the hands of Quantum lct20. And WD200BB again won its favorite set: Visual C++.

Quantum FB Plus AS performed as self-confident as in the previous test, and Seagate Barracuda ATA III yielded its positions a bit (it was probably the disliking to UDMA33 interface that told here). WD200DD won another award in Hi-End test: HE: FrontPage 98 having slightly surpassed Barracuda ATA III. And of course, it proved the best of the best in HE: Visual C++ 5.0!

To illustrate our analysis we would like to offer you a couple of diagrams for Business Disk WinMark and High-End Disk WinMark:

Quantum FB Plus AS and WD200BB hold the lead here, while Seagate Barracuda ATA III falls a bit behind. Why does it happen like that? Shouldn't the higher liner read speed of Seagate Barracuda ATA III tell on the overall performance in these tests? The thing is that when working with smaller files, the fastness of the heads unit means much more for the HDD speed. That's why the hard disk drives tested are now ranged according to their average seek time:

Among the drives with 5400rpm rotation speed, WD Protege (WD200EB) won the race due to larger cache-buffer:

Seagate U5 HDD features slightly bigger average seek time, although the cache buffer is four times smaller than that of WD200EB. As for Fujitsu MPG3204AT, its results are a bit extraordinary: its cache buffer is just the same as that of Seagate U5, but the average seek time is much shorter. Why couldn't Fujitsu beat Seagate U5? Maybe this HDD features less efficient cache buffer algorithms than the Seagate drive...

Here the time required for heads positioning as well as the liner read rate are most important. Before the tests start, the hard disk drives are defragmented and the files are read also linearly.

Since we have already made all the basic conclusions, we won't repeat them here. We would only like to point out that WD200EB showed pretty good results in UDMA33 mode.

Adaptec Threadmark 2.0

Although there has been a lot of criticism expressed towards this test by a number of respected web-sites, it is still one of the best to test the performance of hard disk drives when they work with streaming data (read and write). That is why we would like to offer you the results obtained in this test for all our participants:

Here WD drives, which we have always blamed for low performance, won the lead (among HDDs with 7200rpm and 5400rpm) and proved robust against interface changing. We consider it worth mentioning that Fujitsu and Seagate U5 performed very strangely in UDMA33 mode, i.e. their performance was too low compared to UDMA66.

Although the parameters of Quantum FB Plus AS and WD200BB are very close, the latter managed to prove faster here. Note that there is no correlation with liner read rates of these HDDs: Quantum FB Plus AS can boast higher liner read speed in the beginning of the disk.

WinBench99 for Windows2000 (FAT32)

Here we used the same WinBench99 1.2 but with a different operation system. As the preliminary tests showed, the results appear a bit higher than in case of Win98. That's why we think these results could be of interest to those of you who prefer Win2000.

In Business Disk WinMark there were three winners: Quantum AS, WD200BB and Seagate Barracuda ATA III. They ran neck and neck with each other (we neglect the 50KB/sec difference here). And in High-End Disk WinMark WD200BB fell a bit behind them.

In HE: FrontPage 98 and HE: Photoshop 4.0 WD200EB and Seagate U5 worked wonders and managed to surpass their elder brothers.

Quantum Fireball Plus AS suddenly yielded to WD200BB, which took the lead in this race. Seagate Barracuda ATA III proved really fast and appeared the best in Business Disk WinMark.

As for the drives with 5400rpm rotation speed, they are hardly struggling with each other. WD200EB leaves all the rest really far behind everywhere (it owes its victory to larger cache, no doubt):

Quantum Fireball Plus AS HDD managed to recover a little bit due to good results shown in UDMA33 (the performance drop was insignificant there) and due to the failure of Seagate Barracuda ATA III. WD200BB won in three of the seven Hi-End tests and got the highest final points, as a result.

And now the diagrams, as usual:

It is really impressive, how close WD200EB is to HDDs with 7200rpm!

By some mysterious chance, WD200EB managed to surpass WD200BB in UDMA100 mode, however in UDMA33 and UDMA66 everything returned to its natural course.

WinBench99 for Windows2000 (NTFS)

In this review we decided not to discuss the differences between NTFS and FAT32 file systems, because it should be a separate article. The only thing we would like to mention in this respect is the fact that the default cluster in NTFS makes 4KB. We have already shown when talking about FAT32 that smaller cluster negatively tells on the disk subsystem performance. Will this rule be valid for NTFS as well? The answer is most likely to be "yes". We'll see whether it's true or not. First let's look at the performance of our testing participants in case of default formatting:

WD200BB performed just excellently: small cluster seems to be a beautiful match for cache buffer algorithms of WD Caviar.

The picture is just the same as in the previous case here. WD200BB is an indisputable leader everywhere.

In UDMA33 interface WD and Seagate HDDs feel not quite at home, so Quantum FB Plus AS managed to get a few points.

Seagate hard disk drives simply hate UDMA33 mode so greatly that U5 lost in both tests even to a slower Quantum lct20.

And now take a look at the diagrams:

The winner here is WD200BB, then comes Quantum FB Plus AS, which showed stable performance and didn't face any problems in UDMA33 mode.

HDTach 2.61

One of the peculiarities of this benchmark is the fact that it tests the drives on the physical level, which makes the results independent of the cluster size and type of operation system. This gives us a perfect idea what the hard disk drives are capable of in this particular system. As usual, we suggest studying the tables first:

Graphs	Graph	Graph	Graph	Graph	Graph	Graph	Graph

In terms of "physics", Seagate Barracuda ATA III appears an indisputable leader. We wish the test could show the Read Burst Speed above the 80MB/sec limit. Taking this parameter as a basis, we could single out the following three HDDs: Quantum lct20, Seagate U5 and Fujitsu MPG. Read Burst Speed of these models is much lower than the possible data transfer rate of the controller. What limits the Read Burst Speed parameter then? Since there is one and the same data block requested, cache buffer can't be this bottleneck. So, it is the HDD's DSP that limits the data transfer rate. Since these hard drives aren't intended for showing any performance records, they are equipped with the DSP, which provides the most well-balanced price-to-performance ratio...

Graphs	Graph	Graph	Graph	Graph	Graph	Graph	Graph

Well, the Read Burst Speed theory seems to have fallen apart here :-)

The Read Burst Speed values of our three heroes are again lower than those of the leading ones, but higher than those obtained in UDMA100 mode. It means that these HDDs (to be more exact, their firmware) simply do not like UDMA100 interface (aren't optimized for it, to put it differently)…

Since not all PC owners possess new mainboards or ATA/66 -ATA/100 controllers, we decided to look at the performance in UDMA33 mode as well.

Graphs	Graph	Graph	Graph	Graph	Graph	Graph	Graph

Again Quantum FB Plus AS proves the best in UDMA33 mode. Now let's take a look at the diagrams. Here is what we got:

The good thing about this parameter is the fact that it shows the average read speed on the entire platter and not just the read speed in the first sector with the highest data density.

The indisputable leader here (just like in ZD Disk Inspection Test) appears Seagate Barracuda ATA III. The second place belongs to Quantum FB Plus AS.

Seagate Barracuda ATA III is again ahead of all, however this time the gap between the leader and its closest competitor, Quantum FB Plus AS, got considerably smaller. WD200BB hard disk drive, which was only the fifth in the previous test, here managed to win the second position!

To tell the truth, we were surprised to see so different results obtained for HDDs with the same data density per platter. As is known, data density is not homogeneous and gets lower when moving from the outer to the inner platter edge. The closer we are to the center, the shorter is the physical track that's why the data should be written more frequently in order to fit there the same amount of sectors as on the outer platter edge. The currently used technologies do not allow increasing the data density over a certain limit therefore the number of sectors gets reduced. Some manufacturers, such as Fujitsu, mention the max and min number of sectors per track: 798/441. As you see, the data density drops nearly twice.

Now let see what the disk transfer rate tests show:

Of course, a lot depends on the size and the amount of sectors with the same data density. Take for instance, WD200BB. Its first three sectors feature almost the same data density and can be (hypothetically) regarded as one "longer" sector:

Definitely, this adds to performance of the drive, but will it tell on the reliability of the device? Judging by the graph, the tracks in the second third of the HDDs receive the same amount of data as those in the very beginning of the disk drive. There are two explanations possible. WD has either restricted the number of sectors on the platter outer edge or overloaded the inner tracks.

Certainly, we can't say that the smaller ratio between the disk transfer rate in the beginning of the drive and that in the end is the worst. However, we are quite concerned about really low values of this parameter. Here is what we got after sorting all HDDs in the descending order:

These were just our own ideas, however, you wouldn't deny that different manufacturers evidently have different approaches towards data density distribution.

Intel IOMeter

If you want to know more details about the testing methods of Intel IOMeter, you can take a look here. In general, the main idea of this test is based on a certain model imitating the work of applications in an operation system. This model starts attacking the HDD with the whole bunch of I/O requests. There can be from 1 to 256 simultaneous I/O requests coming, which are called "loads". There are five fixed loads selected.

Fantastic! Quantum FB Plus AS is far ahead of all! And who is its rival? No, it's not Seagate Barracuda ATA III, it's Seagate U5. Well, it's quite understandable why Quantum AS won the race: it has the smallest average seek time and it copes with random requests much better than its competitors. However, as for poor performance of WD200BB, we are at a loss...

The higher gets the amount of I/O requests, the faster proves Quantum AS HDD, because it boasts the best average access time of all the hard disk drives tested.

It is also interesting to observe the struggle between Seagate drives. Sometimes, U5 manages to beat its elder brother even.

Heat and Noise

We measured the heating with the external thermal diode connected to the mainboard. IOMeter takes over 2.5 hours to run just once and the HDD works quite tensely all this time that's why we may regard it as normal thermal conditions for the drive. Although, we didn't put the HDD into a PC case, it was just lying on the table. The room temperature was around 22 degrees Centigrade. As soon as the test came to an end we restarted the system and took the temperature from the system BIOS (1007A):

All drives with lower rotation speeds showed very close results. Only Fujitsu MPG3204 appeared one degree hotter than the others :-)

The drives with 7200rpm speed, were somewhat warmer however, they didn't exceed the dangerous limit. Although you should bear in mind that the HDDs in our tests featured only 1 platter each. So, you should remember that any additional platter adds 10 degrees to the HDD temperature. So, if you are planning to get a 40GB+ HDD then a proper cooling solution will be just what you need to ensure safe and reliable operation.

Unfortunately, we didn't have the device for noise level measuring, that's why we will simply describe our impressions of each particular HDD.

Quantum Fireball Plus AS: This HDD reminded us of the IBM DTLA in terms of noise spectrum. However, the spindle rotates even quieter. It is probably due to the HDB-bearings used in the drive construction. We would give this HDD 4.5 points out of 5.

Quantum Fireball lct20: We didn't hear any spindle noise, however, the heads positioning was carried out louder than in the previous case. We would reward this drive with 4.5 points out of 5.

Seagate Barracuda ATA III: Unlike Barracuda ATA II, this HDD is much quieter, however, you should bear in mind that the 20GB model is equipped with FDB-bearings. You hardly hear any rotation noises, but the heads are positioned quite loudly. We would give this model 4 points.

Seagate U5: This HDD is packed into a special SeaShield that's why when it is not installed into the PC case all vibrations are absorbed by this shield completely. In other words, we didn't hear anything at all. In this case we would give this drive all 5 points, although inside the PC case the heads positioning won't be that noiseless any more.

WD Caviar WD200BB: This is the noisiest hard disk drive of all in this group. The noise made by the rotating spindle is not that annoying, actually, however, the heads cause some really unpleasant sounds. The vibrations during the HDD operation were also quite tangible. Our mark would be 3.5 points out of 5.

WD Protege WD200EB: Strange as it might seem, but although this hard disk features similar construction to the previous one, the noise level is impressively low. Only because of the noticeable vibration, this HDD deserved 4 points.

Fujitsu MPG3204AT: It looks as if we were simply unlucky with the unit, because this hard disk rotated with such a noise as if it were not a 5400rpm model, but a one-year-old 7200rpm one. The heads unit worked fine. Therefore, the total points made 4.

Conclusion

Having calculated all the points in the tests described above, we got the following situation. The first three awards received:

As for the first prize, we don't have any doubts about it, because here our subjective impressions were fully backed up by the testing results. Quantum Fireball Plus AS deserved the first prize for high results shown in Business Disk WinBench and excellent performance in IOMeter. We should also mention low average seek time and no problems in UDMA33 mode. This hard disk drive is the quietest of all and dissipates less heat.

The second prize is given to WD200BB. Although it proved not very well in Business Disk WinBench, it managed to regain its reputation in Win2000 + NTFS.

Despite really impressive muscles, Seagate Barracuda ATA III won only the third prize. This HDD performed not very well in UDMA33 mode and IOMeter. In file tests it looked better, though failed to beat WD200BB. The only thing that justifies it is the fact that we had only a sample, and not a mass piece. Maybe with new firmware this hard disk appears better.

In the second group the winners were: WD200EB, Seagate U5 and Fujitsu MPG3204AT.

WD200EB features a four times larger cache buffer than its competitors that's why it was the leader in all file tests.

Seagate U5 performed stably and sometimes tried to compete with Seagate Barracuda ATA III (especially in synthetic benchmarks).

To tell the truth, Fujitsu MPG3204AT disappointed us a bit. It showed good average seek time, however, in all the tests occupied the third place only. Well, not bad for a working horse…

Quantum Fireball lct20 managed to run all the tests up to the end and will be rewarded with our special prize: "For Persistence" :-)