by Aleksey Meyev
08/15/2009 | 09:41 AM
Some three months ago we welcomed early hard disk drives based on 500GB platters but now the situation is completely different. 1.5- and 2-terabyte models from Samsung, Seagate and Western Digital are all available for purchase today, even though all of them belong to power-efficient series with reduced spindle rotation speed. At the time of our writing this, Seagate’s Barracuda 7200.12 is the only series that has a spindle speed of 7200rpm and 500GB platters, but its maximum capacity is 1TB. If you want more storage space, you have to put up with the lower spindle speed. Such HDDs can actually make a very good second hard disk for storing data only. You don’t need high performance from such a disk whereas the lower power consumption and heat dissipation would come in handy. This lucky combination of large capacity and low heat dissipation makes these HDDs suitable for modestly loaded file servers and for household electronics (it is always better to have the opportunity to record 2 terabytes of movies instead of 1 terabyte).
If you are only into 7200rpm models and want to have the highest storage capacity possible, you should just wait a little till the 2-terabyte Western Digital Caviar Black and the similar product from Hitachi hit the shops. By the way, Hitachi is the only maker that begins to produce HDDs larger than 1TB from full-speed models. Hitachi doesn’t offer a power-efficient series. So, until we’ve got large and fast HDDs, it is just the proper time to have a comparative test session and see what today’s high-capacity HDDs can offer.
They are actually not so very slow, those modern 5400rpm drives. Of course, they will always be inferior to HDDs with higher spindle rotation speed, but right now they have an advantage in the way of larger-capacity platters. Although these platters do not provide high sequential speeds, the high recording density makes up for the higher seek time. Here’s the clarification: the average seek time is measured for the full capacity of an HDD but it is usually only a tiny portion of the capacity that is being used at the given moment. If data on your hard disk are not too fragmented, all of a specific application’s data will be located within a few cylinders’ space. The width of a cylinder is somewhat smaller with HDDs that have a larger number of platters and a higher track density (as we already know, this is what modern 500GB platters are in comparison with 333GB platters). So, if we compare a 3-platter 1TB and a 3-platter 1.5TB drive, the latter is going to have only two thirds of the physical cylinder width of the former due to the higher track density. And the 25GB or something zone with your OS and frequently used applications on the 1.5TB drive is going to be only two thirds as wide as on the 1TB drive. It means that, if the heads move at the same speed from one cylinder to another, data are accessed faster on 500GB platters.
You are going to see both 1.5TB and 2TB drives in this test session. There are not enough HDD models with a capacity of over 1 terabyte for two reviews. By the way, the cost of storage per gigabyte is considerably higher with the 2TB models. This should be taken into account when you go shopping for your new hard disk.
First goes a model from Samsung’s new EcoGreen F2 series which was one of the first series with 500GB platters to reach the market. The company’s previous economical series was released after the standard F1 DT line-up, but now Samsung follows the suit and, like the other makers, offers 5400rpm models first. However, these HDDs have all the characteristic features of the previous generations of Samsung’s products such as flight height control, a dual-core processor and a vibration sensor. This is in fact the obligatory minimum which is available, under different names, in products from all HDD makers.
As opposed to the others, Samsung does not produce 4-platter models, so 1.5 terabytes is the maximum capacity available. Besides the model with 32MB cache (this is the sample we have got for our tests), the EcoGreen F2 series includes its counterpart with 16 megabytes of cache.
The rather old Seagate Barracuda 7200.11 is the only model in this review to have a spindle rotation speed of 7200rpm. This HDD belongs to a transitional generation, actually. Its capacity of 1.5 terabytes is achieved by means of four rather than three platters (so the recording density is 375 rather than 500GB per platter). It is a kind of a reference point that we will compare the other models with.
Like the other manufacturers, Seagate once used to produce 5400rpm drives, e.g. Seagate U6, to achieve a lower manufacturing cost and higher reliability together with reduced noise. Then, the cost benefits vanished, the reliability increased, the bearings got quieter, and 5400rpm HDDs left the market. The manufacturers found it more economically profitable to release extensive 7200rpm HDD series including small-capacity models. But now, following the other makers, Seagate has begun to produce HDDs with lower spindle speed again.
To be exact, the spindle rotation speed of the new Barracuda LP series is not 5400 but 5900rpm. There is nothing extraordinary about that. The traditional speeds of 4200-5400-7200rpm simply represent round numbers, i.e. 70-90-120 rotations per second. Moreover, the spindle speed is not something fixed. It varies from one specific sample of a HDD to another (usually within a 20rpm range). Long ago, you could even see a HDD with a specified spindle rotation speed of 3524rpm, for example. We suspect that Seagate had to use this speed for the spindle because its hardware platform could not be used to make HDD with high recording density, more than two platters, and with a spindle rotation speed of 7200rpm. Therefore the speed is reduced to 5900rpm. The other parameters being equal, this gives the HDD an edge against its opponents in terms of sequential speed and access time. And of course, this is a clear marketing advantage. Thus, Seagate has become competitive as it now offers economical HDDs of large capacities. By the way, the letters LP stand for Low Power rather than Low Profile. The latter term is applied to expansion cards.
The 2TB Western Digital was the first (and, for some time, the only) hard disk drive of that capacity that came to retail. It was covered in our previous review, but has got much more opponents today.
The 1.5TB model from Western Digital differs from its larger cousin with a considerably lower price (but the Samsung is even cheaper as yet). Like the 2TB model, this HDD comes with 32MB of cache only.
If you take a look at the official specs, you can note that this model weighs as much as the 2TB model. So, it is quite possible that this HDD is based not on three full-featured platters but on four reduced-capacity platters (or has disabled heads). This would be a reasonable solution. If there are few 500GB platters as yet and the product yield is rather low, why not format the rest of the platters for a lower capacity? Considering this drive’s IOMark results, we write four platters and eight heads into its specs. We can change our opinion if we see any evidence to the contrary.
Western Digital tried to make the most from its early coming to the 2TB sector. So, after the desktop series, the AV-GP series was expanded as well. The key feature of this series is special firmware which is optimized for simultaneous delivery of up to 12 HD-video streams. Some time ago we tested a special-purpose disk (that was the 1TB Seagate SV35.3) and its performance did not impress us at all. Perhaps this one will do better.
Western Digital claims the AV-GP series to use special PWL technology (Preemptive Wear Leveling). Its point is that the HDD moves its heads above the entire platters from time to time. The company claims this helps reduce the nonuniform wear of the platters typical of HDDs that are employed for video applications. Unfortunately, Western Digital does not reveal all the details and we don’t really know how the wear of the platters can be affected by the actuator with heads sweeping over the platter surface. Perhaps the heads are working at that, reading and writing data back?
There is one more Western Digital drive in this review. The company has introduced a new series of power-efficient HDDs for RAID arrays that is called RE4-GP. The series consists of only one, 2TB model so far, but it is interesting anyway.
Besides the typical technologies of Western Digital’s server-oriented HDDs, this model features RAFF (Rotary Acceleration Feed Forward) which helps minimize the influence of vibration produced by multiple HDDs in a rack. You can refer to the related document for details, but we want to note one curious fact. The influence of vibration used to be only accounted for in high-speed Raptor HDDs with their 10,000rpm, but now vibration is a factor even for HDDs with half that spindle speed. With the track density being so high now, even a weak impact can beat the head off the necessary track.
The second feature of the new series is the cache buffer which is now as large as 64 megabytes. Yes, this is the next milestone in the growth of HDD cache. From 2 megabytes HDDs have progressed to 8 and 16 megabytes of cache. Today, 32 megabytes is the most common size. And now, we’ve got a HDD with 64 megabytes of cache memory. On one hand, you can’t have too much of cache. But on the other hand, this cache must be used effectively. It won’t do if looking for data in the cache takes too much time. We will be able to see the benefits of the bigger cache by comparing this HDD with the very similar WD20EADS.
The following table lists the specifications and firmware versions of the tested HDDs.
Click to enlarge
Take note that we tested the Seagate Barracuda 7200.11 with the potentially dangerous firmware version. When we try to update it, they say that, according to the serial number, our HDD is problem-free. And it confirms this by having worked without any problems so far.
The following testing utilities were used:
We installed the generic OS drivers for the drives and formatted them in FAT32 and NTFS as one partition with the default cluster size. For some tests 32GB partitions were created on the drives and formatted in FAT32 and NTFS with the default cluster size, too. In every test, save for IOMark and the Defragmentation test, the drives were connected to a Promise SATA300 TX4302 controller installed into a PCI-X slot and had their NCQ support enabled.
We will be using abbreviated names of the HDDs to make the review more readable.
We use our internal IOMark tool for low-level tests. Let’s check out the sequential read speed of the drives first.
Now we can compare the HDDs by the read speed at the beginning and end of the partitions created on them.
The Seagate 7200.11 is the leader of this test thanks to its 7200rpm. The second-best Seagate LP is very good, too. Its data-transfer rate of 120MBps is very high, especially for a HDD with lower spindle rotation speed. Does it benefit from having an extra 500rpm in comparison with the other power-efficient HDDs? Yes, but not much. If you take a closer look at the data-transfer graphs, you will see that the speed of each HDD varies wildly depending on how lucky the specific surface/head pair was during formatting. And this fluctuation can easily negate any advantage from the somewhat higher spindle speed.
For all those fluctuations of speed, one graph is especially interesting as it differs greatly from the others. It is the graph of the 1.5TB model from Western Digital. With full-size platters, the speed at the beginning of a disk is roughly twice as high as at the end. Here, we see that the speed at the end is obviously too high. This seems to be the consequence of the HDD’s using only the fastest zones of platters which are not formatted for their full capacity.
Now, let’s check out the speed of the HDDs’ cache memory.
Western Digital’s HDDs all behave in a similar manner when working with the cache buffer. They are all somewhat slower than their opponents in terms of top speed. But the main differences are in the graphs. That is, in the speed of processing large data chunks. The Samsung looks best here. It has no slumps and its write speed is not much slower than its read speed: the graph is very neat.
Western Digital’s products are somewhat worse in this respect. Their writing is slower than their reading and subject to fluctuations. The RE4-GP stands out among this company’s products. Its speed slumps on very large data blocks. The new firmware seems to be not quite optimized for the enlarged cache.
Seagate’s HDDs are far from good, too. The 7200.11 suffers a performance hit on data chunks larger than 256 sectors (128KB): the read speed plummets to 150MBps and the write speed, to 100MBps. As a result, the very purpose of the cache buffer is negated because the speed is comparable to the speed of sheer writing to platters. The new LP is better in this respect. Its speed does not fluctuate too much. Its read speed does not sink below 200MBps and its write speed is a decent 150MBps. Still, the example of the Samsung drive proves that firmware can be improved even more.
From the low-level IOMark to the synthetic IOMeter. IOMeter is sending a stream of read and write requests with a request queue depth of 4. The size of the requested data block is changed each minute, so that we could see the dependence of the drive’s sequential read/write speed on the size of the data block. This test is indicative of the maximum speed the drive can achieve.
The numeric data can be viewed in tables. We will discuss graphs and diagrams.
The Seagate 7200.11 boasts the highest speed in this test, too. However, on small data chunks it is inferior to all the new HDDs, save for the Samsung which accelerates to its top speed rather sluggishly. The Samsung’s top speed is somewhat higher than that of three models from Western Digital, though. The fourth Western Digital, the RE4-GP model, is faster than its cousins throughout the test.
The Seagate LP is the best of the power-efficient models. Its extra 500rpm seem to show up here.
We don’t see anything particularly new when we switch from sequential reading to sequential writing. We’ve got the same two drives from Seagate as leaders, and the Samsung is again slow on small data chunks.
In this test IOMeter is sending a stream of requests to read and write 512-byte data blocks with a request queue of 1 for 10 minutes. The total number of requests processed by the HDD is over 60 thousand, so we get a sustained response time that doesn’t depend on the HDD’s buffer size.
The Western Digital RE4-GP is ahead at reading. Western Digital squeezes everything out of the heads of its server series. Those 14 milliseconds is an excellent response time for a 5400rpm drive. Second place goes to the Seagate LP which is 0.5 milliseconds worse than the leader. The 2TB WD Caviar Green is good, too. It enters the top three and even beats the Seagate 7200.11, despite the latter’s higher spindle rotation speed. The 1.5TB model and the video-optimized product deliver modest results here. We suspect that the AV-GP has somewhat slowed-down heads. It is the heads that produce the most noise whereas home electronic appliances are expected to be quiet.
We’ve got the same leader at writing: the HDD from Western Digital seems to make good use of its 64 megabytes of cache. Its response time is twice better than at reading. Among the other results we can note the high response time of the Seagate 7200.11. As we know from earlier tests, this drive’s electronics gets “stifled” when processing small-size data chunks.
In the next test the drive is being bombarded with read requests like in the response time test, and we calculate the difference between the LBA addresses of the previous and next requests and divide it by the time it took to perform the request. In other words, we have the distance (in gigabytes) the drive can run through in 1 second. The results are averaged and compared.
The 2TB models are superior here. We have found out in the previous test that the Western Digital AV-GP has slowed-down heads but even this model is better than the 1.5TB ones. Take note that it is not high-density platters (the Samsung has them, too) and not the 4-platter design (the Seagate 7200.11 uses it as well) but the combination of the two that is the decisive factor here. A low response time – as the consequence of quick heads – would be appropriate, too. It helps the Western Digital RE4-GP win this test.
Now we’ll see the dependence between the drives’ performance in random read and write modes on the size of the data block size. We will only discuss the processing of small data blocks measured in operations per second. With large data blocks, the performance depends on the drive’s sequential speeds.
IOMeter: Random Read, operations per second
Reading small data chunks is far from interesting: the HDDs are ranked according to their response time. The lower the response time, the higher the performance is.
IOMeter: Random Read, megabytes per second
The HDDs behave in a similar way with larger data blocks, too. You can only see them differ at 2MB and larger blocks when the sequential speed becomes the decisive factor. As expected, the Seagate HDDs are in the lead then, the 7200.11 being somewhat faster than the LP.
IOMeter: Random Write, operations per second
The huge cache and quick heads ensure a solid advantage to the Western Digital RE4-GP. This HDD easily copes with writing small-size random-address data blocks. The other power-efficient HDDs go neck and neck, the 1.5TB Western Digital being somewhat better and the Samsung somewhat worse than the others.
IOMeter: Random Write, megabytes per second
When the data blocks get larger, the Western Digital RE4-GP loses its advantage and is overtaken by the Seagate 7200.11 that has a higher sequential write speed.
In the Database pattern the drive is processing a stream of requests to read and write 8KB random-address data blocks. The ratio of read to write requests is changing from 0% to 100% with a step of 10% throughout the test while the request queue depth varies from 1 to 256.
You can click this link to view the tabled results for IOMeter: Database pattern.
We will build diagrams for request queue depths of 1, 16 and 256.
Things are quite interesting at the minimum queue depth. The Western Digital RE4-GP is ahead of the other HDDs both at reading (thanks to its quick heads) and writing (thanks to its huge cache). The other graphs are close to each other except that the Samsung is slow at writing and the Western Digital AV-GP and the 1.5TB Caviar Green are slow at reading. And while the results of the AV-GP can be explained by its slow heads, we don’t know what happens to the 1.5TB Caviar Green at pure random reading.
When the queue depth is increased to 16 requests, we have some more surprises. The WD RE4-GP is still beyond competition at writing, but loses its leadership at reading. It is slower at reading than the Seagate HDDs that have higher spindle rotation speed as well as than the 2TB Caviar Green. The 1.5TB Caviar Green is again the worst drive at reading but the Samsung is the overall worst throughout all request queue depths.
The competition gets tougher at high loads. The Western Digital RE4-GP is unrivalled at writing, but finds itself the slowest of all at reading. The Seagate 7200.11 wins at reading, followed by the WD Caviar Green series in which the 1.5TB model goes ahead at pure reading. The Samsung does not like server loads and fails this test.
Winding up this part of our tests, we will build diagrams showing the performance of each HDD at five different queue depths.
The Samsung EcoGreen F2 resembles the desktop F1 DT series with its weak NCQ and modest deferred writing. It also has the specific behavior at high percentages of reads that makes us suspect that its heads are deliberately slowed down.
The Seagate HDDs have everything a decent HDD must have. And it is good that the Barracuda LP at 5900rpm is about as fast as the Barracuda 7200.11. However, the power-efficient model has occasional performance slumps under medium queue depths, so its firmware needs improvement in terms of request reordering.
Western Digital’s Caviar Green drives behave in the same manner, except that the 1.5TB model is slower at pure reading at any queue depth. This must be a feature of its firmware whether intentional or not.
The Western Digital AV-GP is similar to the other HDDs from the same company but behaves in a calmer manner. It has very modest deferred writing and weak NCQ.
And finally, the RE4-GP differs greatly from the other HDDs. It boasts superb deferred writing: it can really put those 64 megabytes of cache memory to good use. However, it has no request reordering whatsoever. This is actually no problem for an HDD which is designed for RAID arrays because the RAID controller can take care of request reordering and the lack of NCQ in the HDD even helps avoid potential conflicts. However, it is rather odd to see an NCQ-less drive in the year of 2009. This is the reason why it is slower than the others at every request queue depth save for the minimum one.
The drives are tested under loads typical of servers and workstations.
The names of the patterns are self-explanatory. The Workstation pattern is used with the full capacity of the drive as well as with a 32GB partition. The request queue is limited to 32 requests in the Workstation pattern.
The results are presented as performance ratings. For the File-Server and Web-Server patterns the performance rating is the average speed of the drive under every load. For the Workstation pattern we use the following formula:
Rating (Workstation) = Total I/O (queue=1)/1 + Total I/O (queue=2)/2 + Total I/O (queue=4)/4 + Total I/O (queue=8)/8 + Total I/O (queue=16)/16.
When there are only read requests to be processed, Seagate HDDs take the lead, the 5900rpm newcomer being but slightly slower than its 7200rpm predecessor. Western Digital’s drives are similar to each other. The WD RE4-GP heads this group at short queue depths due to its rather quick heads, but its performance does not grow up much along with the queue depth because it lacks NCQ. As a result, it is downright slower than its cousins at long queue depths. The Samsung is too weak in comparison with the other HDDs.
When there is even a small share of write requests to be processed, the Western Digital RE4-GP takes the lead thanks to its large cache. But it lacks NCQ and is far from brilliant at long queue depths where the Seagate 7200.11 is supreme. Interestingly, it is not the power-efficient model from Seagate but the 2TB Green model from WD that is competing with the leader here. The Samsung is the worst performer, again.
The Western Digital RE4-GP has the highest overall rating, according to our formula.
When the share of writes is increased more and the queue depth is reduced, the WD RE4-GP wins. Next goes the Seagate 7200.11 which is followed by the power-efficient model from the same maker.
Take note of the 1.5TB Western Digital Caviar Green. It suffers a sudden performance hit at the shortest queue depth. This is the same peculiarity in its behavior as we have seen in the Database pattern.
When the test zone is reduced to 32GB, the HDDs with higher spindle rotation speeds go ahead. For example, the Seagate LP is obviously better than the others but cannot catch up with its predecessor. The WD RE4-GP has the best results among the 5400rpm models, and the Samsung is again slower than the rest of the HDDs.
Summing up this part of our tests, we can say that the WD RE4-GP is very good but it will only show its best in a RAID array. The Seagate LP is going to be better as a single disk. The Samsung is slow under server loads. Hopefully, Samsung’s server-optimized models will be better in this respect.
The multithreaded tests simulate a situation when there are one to four clients accessing the virtual disk at the same time – the clients’ address zones do not overlap. We will discuss diagrams for a request queue of 1 as the most illustrative ones. When the queue is 2 or more requests long, the speed doesn’t depend much on the number of applications. You can also click the following links for the full results:
It is easy to see the trend at multithreaded reading. We’ve got two leaders: Seagate LP and Western Digital RE4-GP. Although the WD AV-GP is claimed to be optimized for multithreaded load, it is nothing but mediocre in this test.
And we must note once again that the current firmware of Samsung’s HDDs cannot identify multithreaded load and slips to random reading mode with consequently poor performance.
Every HDD copes well enough with multithreaded writing. The Samsung and the WD RE4-GP win by a very small margin. The latter seems to make good use of its large cache again because the other three drives from WD have similar and somewhat lower results.
For this test two 32GB partitions are created on the disk and formatted in NTFS and then in FAT32. A file-set is then created, read from the disk, copied within the same partition and copied into another partition. The time taken to perform these operations is measured and the speed of the disk is calculated. The Windows and Programs file-sets consist of a large number of small files whereas the other three patterns (ISO, MP3, and Install) include a few large files each.
We’d like to note that the copying test is indicative of the drive’s behavior under complex load. In fact, the HDD is working with two threads (one for reading and one for writing) when copying files.
This test produces too much data, so we will only discuss the results achieved with the Install, ISO and Programs file-sets in NTFS. You can use the links below to view the full results:
The Western Digital RE4-GP wins the test of creating files, enjoying the biggest advantage with the mixed files of the Install pattern. The Samsung is also good, but its superiority is only obvious with large files.
It is Seagate’s HDDs that are poor at writing files. The 7200rpm model is considerably slower than the opponents and the new Barracuda LP is slower still.
Seagate’s HDDs show their best at reading files. They have top places irrespective of the file-set. The Barracuda LP, being slower than its predecessor with higher spindle speed, is considerably better than the Samsung which is third.
We’ve got two leaders when copying within the same partition. They go neck and neck in the ISO pattern. The server model from Western Digital wins the Install pattern and the Samsung is the best at copying the small files of the Programs pattern.
Seagate’s HDDs are much slower than the others with the large files of the ISO pattern but in the other patterns the gap is very small.
The picture is more vivid when the files are copied from one partition to another. We’ve got the same leaders but the Samsung is rather on the losing side in the Install pattern and does not enjoy a large lead in the Programs pattern. As for Seagate’s HDDs, they are obviously poor with every file-set, the new Barracuda LP being somewhat slower than the older 7200.11.
PCMark 2005 has the same tests as the 2004 version (not only in names, but also in results as we have seen a lot of times), so we only use one test from PCMark 2004 which is not available in the 2005 version. It is called File Copying and measures the speed of copying some set of files. The other tests are:
The final result of the average of ten runs of each test.
PCMark does not always repeat the results of FC-Test, but in this test session they agree that the Western Digital RE4-GP is the best at copying. The Samsung is second best whereas the Seagate HDDs are poor in this test.
It is Seagate’s HDDs that are ahead of the others at booting the OS up, the LP model being somewhat better than its predecessor notwithstanding the lower spindle rotation speed. The other HDDs are similar to each other.
It is the server model from Western Digital that copes best with loading applications. Second place goes to the single 7200rpm HDD in this review. The Barracuda LP is third.
Under the general usage load the Seagate 7200.11 and Western Digital RE4-GP are contending for first while the Seagate LP and Western Digital AV-GP are fighting for third place. The AV-GP is surprisingly faster than its Green series mates.
Scanning for viruses is highly sensitive to caching mechanisms and Seagate’s HDDs behave better here, the new Seagate LP being faster than its predecessor. All the HDDs from Western Digital roll back to last positions, the model with 65MB of cache giving way to the two other 2TB models.
Writing files in PCMark is obviously different from what we have seen in FC-Test. The Seagate 7200.11 takes first place, without showing any sign of a performance slump. Second place goes to the Samsung. It overtakes the Seagate LP which has accelerated, too.
According to the 2005 version of the benchmark, the Seagate 7200.11 has the best overall score. The Barracuda LP is just a little bit worse. The Western Digital RE4-GP is third in this test.
To make this part of our test session complete, we are going to run the latest version of PCMark called Vantage. Compared with the previous versions, the benchmark has become more up-to-date and advanced in its selection of subtests as well as Windows Vista orientation. Each subtest is run ten times and the results of the ten runs are averaged.
Here is a brief description of each subtest:
Basing on these subtests, the drive’s overall performance rating is calculated.
The two HDDs from Seagate and the Western Digital RE4-GP show the best performance in this test that emulates multithreaded load. The WD AV-GP, allegedly optimized for multithreaded load, does not show anything exceptional again.
The same three HDDs are going to be appreciated by gamers. The Samsung is a disappointment in this test. It is far too slow.
We see the two leaders of this test session – Western Digital RE4-GP and Seagate Barracuda LP – take top places again when loading images into a photo gallery. The Samsung is a loser again.
It is the 7200rpm model that boots Windows Vista faster than the others. Next go the WD RE4-GP and Seagate LP. The Samsung finds itself in last place again. It looks like processing files is the only thing this HDD is really good at.
There are some changes in the leading trio in the Movie Maker test. The Seagate 7200.11 drops out of it and falls to bottom place while the Western Digital AV-GP is second. Optimized for home video appliances, this HDD seem to like to work with movies.
This test is sensitive to caching, and Seagate’s products are usually superior in it. This time around, Seagate’s HDDs are good, but the Samsung is even better. Its caching algorithms prove to be stronger!
We’ve got our traditional trio of leaders: the two models from Seagate and the server model from Western Digital. The other models, save for the Samsung, are but slightly slower, though.
And we see the same three leaders again in the Application Loading test. Take note that the best of the new-generation HDDs, for all their power efficiency, are once again competitive to the Seagate 7200.11.
The overall scores are interesting enough. The power efficient Seagate LP has the highest score, being barely ahead of the Western Digital RE4-GP. Third place goes to the previous-generation Seagate 7200.11 that has a higher spindle rotation speed than the others. Interestingly, the WD AV-GP is ahead of the three other power-efficient models (the two WD Caviar Green HDDs and the Samsung EcoGreen) which take last places with almost identical scores.
Next goes our homemade test of defragmentation speed. We created a very defragmented file system on a 32GB partition of a hard disk by loading it with music, video, games and applications. Then we saved a per-sector copy of the disk and now copy it to the HDD we want to test. The tested HDD is connected to the mainboard’s SATA controller whose operation mode (AHCI/Standard SATA) is controlled from the mainboard’s BIOS. Next we run a script that evokes the console version of the Perfect Disk 8.0 defragmenter and marks the time of the beginning and end of the defragmentation process. Thus, each drive is tested twice – with AHCI support turned on and off on the controller. You can refer to this article for details about this test.
The Western Digital RE4-GP wins this test, but by a very narrow margin. Its higher recording density and quick heads seem to do the trick. On the losing side there are two drives from Seagate (the Barracuda LP is depressingly poor) and the WD Caviar Green which seems to have problems with AHCI support.
Now we are going to introduce to you one more interesting test in which we use WinRAR version 3.8 to compress and then uncompress a 1.13GB folder with 8118 files in 671 subfolders. The files are documents and images in various formats. These operations are done on the tested HDD. WinRAR is usually used to benchmark CPUs, but it can make a good test for HDDs if you select the lowest compression level and use a huge amount of files. The HDD’s performance should affect the speed of compressing/uncompressing then.
We tried this test with 2.5-inch HDDs and now will apply it to 3.5-inch ones.
The HDDs are surprisingly similar in the compression test: five out of the seven HDDs turn in nearly identical results. Two HDDs have fallen behind: these are the Western Digital RE4-GP and Seagate LP which have been rather good in the previous tests. The gap is small, though. This test even at our slack settings seems to be highly sensitive to the CPU’s speed and almost indifferent to HDD performance.
Uncompressing data is quite a different story. Every HDD is different, the gap between the leader and outsider amounting to 20%. First place is taken by the Western Digital RE4-GP which seems to make good use of its large cache when writing the numerous small files. Last places are occupied by the Barracuda LP and the WD AV-GP.
You can refer to our article called Hard Disk Drive Power Consumption Measurements: X-bit’s Methodology In depth for details on this test. We will just list the specific modes we measure the power consumption in:
Let’s check out each mode one by one.
The electronics of each HDD consumes about the same about of power from the 5V line when starting up. The WD RE4-GP with 64MB of buffer memory needs somewhat more than the others, though. The Seagate HDDs consume much more power from the 12V line than the others. Hopefully, they need such a high current only to speed the platters up faster.
When idle, the HDDs with lower spindle speed prove to be economical indeed. Even the worst of them is 2 watts better than the Seagate 7200.11. The Seagate LP must be noted again. Despite the somewhat higher rotation speed (in comparison with the other power-efficient models), it is in the top three in power efficiency, being but slightly worse than the two HDDs from Western Digital. We have only one question left: what are the other two HDDs from Western Digital doing if their 5V power consumption is two times as high as that of the other power-efficient HDDs?
The electronics don’t do much at random reading. It is the HDD’s mechanics that does most of the job. The 7200rpm model needs 1.5W of power more than the others and is proportionally hotter. The HDDs with slowed-down heads – the Western Digital AV-GP and the 1.5TB Caviar Green – have somewhat better results than the others.
Deferred writing algorithms come into play here. Consequently, the Seagate HDDs are better than the others in terms of 5V consumption, but worse in terms of 12V consumption. The Samsung is the overall winner. The WD RE4-GP is the worst drive here.
At sequential reading it is on the +12V line that we can see most differences. There are no winners but there are losers. There is one loser, actually. It is the old Seagate 7200.11 with high spindle rotation speed. It consumes 1.5W more than its opponents.
The sequential writing results are similar to those of the previous test. The HDD with higher spindle rotation speed loses again. We must note a couple of facts, though. The WD Caviar Green saves on the 12V line but consumes more than the others on the 5V line. The Seagate LP is good. Although it has a somewhat higher spindle rotation speed than the other power-efficient HDDs, its power consumption is more than modest.
It is the Western Digital RE4-GP that has showed the most memorable performance in this test session. The combination of fast heads, huge 64MB buffer (which is made a very good use of) and optimized firmware make it a leader under server loads. It is in RAID arrays that it is going to perform best. When single, it will suffer through the lack of NCQ support.
The Seagate Barracuda LP is a good product, too. Seagate has made a good debut on the market of power efficient hard disk drives. There is nothing extraordinary about this HDD but it represents a lucky compromise of performance, noisiness and power consumption.
Samsung’s EcoGreen2 is no record-breaker as the main disk of a workstation and is far from brilliant under server loads, but it was confident and occasionally better than the others at processing files. Thus, we can recommend it as a hard disk for storing large amounts of data on a home PC.
The Western Digital AV-GP is an odd product. This model is meant for home electronic appliances but does not have any distinguishing features save for somewhat slowed-down heads.
As for the two Green series models from Western Digital, the 1.5TB one is somewhat sluggish. On the other hand, the 2TB model also looks slow in comparison with its server cousin that has a larger cache and faster heads.
So, we have got a general picture of today’s power-efficient HDDs. They have not become as much faster after the transition to denser platters as the most optimistic buyers might wish. These are just HDDs of high storage capacity with a very good combination of low power consumption and high enough speed. Now we are waiting for the next generation of HDDs to see if they can surprise us in any other way, besides larger capacity.