2.5-Inch SAS Hard Disk Drives Roundup

For a few years we have been putting aside an absolutely new type of hard disk drives: server solutions in 2.5-inch form-factor. We realized it was a big omission on our part and today we are ready to rehabilitate ourselves at least partially by introducing to you six solutions like that.

by Aleksey Meyev Nikita Nikolaichev
12/14/2009 | 02:56 PM

For a few years we have been ignoring a completely new class of hard disk drives – server disks of the 2.5-inch form-factor. It’s time to improve the situation now.

 

First of all, let’s recall why this product class was invented. Quite a long time ago, when we were all much younger and the recording density was much lower, when SSDs were not even heard of and CPUs had legs, some clever engineer must have recalled the famous maxim that God is always on the side of the large battalions. So, that engineer came up with an idea of how to improve a disk subsystem performance by means of adding more disks into it. The idea is quite simple and, judging by our reviews of RAID controllers, effective. If the controller is capable enough, the performance of the disk subsystem scales up well depending on the number of disks. However, a standard HDD rack can only accommodate a limited amount of disks even when you want to add more and more to boost your disk performance even higher. The solution is as simple as the RAID concept itself: the physical size of a hard disk drive must be reduced!

Indeed, if our priority is the number of input-output operations per second rather than the storage capacity, it makes sense to replace physically larger HDDs with smaller ones. For example, you can install six 2.5-inch drives into a 1U server instead of four 3.5-inch ones – an advantage of two spindles! The only question is whether the smaller HDDs have high speed characteristics. Of course, they are going to have lower sequential speeds, but what about the number of disk operations per second?

Testing Participants

Fujitsu MBB2 RC: MBB2147RC – 147GB

 

Like our previous roundup of SAS drives, this review begins with Fujitsu which has sold its HDD manufacture to Toshiba. And we again want to express our wish that the new owner will go on developing this product line. Right now, there are four brands offering such products because Western Digital has just recently announced its S25 series (2.5-inch drives with 10,000rpm and SAS).

Take note that this HDD, like all of its opponents, has a thickness of 12.5 millimeters instead of 9.5 millimeters as it typical of ordinary 2.5-inch drives. When it comes to slower 2.5-inch drives, a thicker case means that the drive has a third platter (and a third pair of read/write heads), but here the thickness is determined by the advanced heads actuator. For SAS drives, even such compact ones, access time is a highly important characteristic and must be improved by any means possible. Therefore they have a high spindle rotation speed as well as fast heads which require a robust actuator.

Unfortunately, we could not get the newest models of the MBD2 RC (with 150GB platters) and MBE2 RC (with 73.5GB platters but with a spindle speed of 15,000rpm) series, so we can only test the previous MBB2 RC series with 73.5GB platters and a spindle speed of 10,000rpm. This series is not the oldest, though. Fujitsu’s first 2.5-inch SAS drives had only half that storage capacity.

Of course, the capacity of this drive does not look large in comparison with newest 3.5-inch SAS or modern 2.5-inch 5400rpm products, but reliability and compactness are often more important for server applications than record-breaking specs. Some people may deliberately choose previous-generation products just because they don’t need highest specs possible but want to save some money. The MBB2 RC series has all the characteristic and demanded traits of 2.5-inch SAS drives such as compactness, low access time (we’ll check this out in practical tests shortly) and low power consumption.

Hitachi Ultrastar C10K147: HUC101473CSS300 – 147GB

 

Hitachi entered the market of compact SAS drives later than the other brands. It did not have products with 36.5GB platters but started off with twice larger platters instead. The C10K147 series was the company’s first attempt at making 2.5-inch drives with a spindle rotation speed of 10,000rpm. Despite the rather small capacity of the platters, the series features all server-oriented technologies such as the parking of the heads on a ramp, thermal flying height control, and Rotation Vibration Safeguard.

Hitachi Ultrastar C10K300: HUC103030CSS600 – 300GB

 

Hitachi has recently updated its compact SAS products by releasing a new series. And this time the company is one of the first makers to do so. The C10K300 series uses twice larger platters (the dual-platter model is now 300GB large) and a SAS 6Gbps interface. An individual disk does not need such a high interface bandwidth (even the best of 15,000rpm drives are only as fast as 200MBps) but the new high-speed interface is going to be useful for connecting disk racks.

We don’t have a Hitachi С15K147 model in this review, but want to say a few words about it. Released along with the C10K300, it has a SAS 6Gbps interface, too, but does not have such large platters. Instead, it has a spindle rotation speed of 15,000rpm, which is an excellent option for users who want to get as many operations per second from their disk array as possible.

Seagate Savvio 10K.1: ST973401SS – 73GB

 

It’s a kind of a rule that no HDD review can be done without Seagate’s products. After all, it was the Seagate Savvio 10K.1 series that first implemented a spindle speed of 10,000rpm and a SAS interface in the 2.5-inch form-factor. The recording density used to be very low then. This 73GB model has two platters to deliver its storage capacity.

Seagate Savvio 10K.2: ST973402SS – 73GB

 

After a while, Seagate introduced the Savvio 10K.2 series with double recording density. The senior model now had a capacity of 146GB.

Seagate Savvio 15K.1: ST973451SS – 73GB

 

At about the same time as the 10K.2, Seagate released its first series of compact HDDs with a spindle rotation speed of 15,000rpm which was appropriately named Savvio 15K.1. Of course, the increased speed of the spindle did not come for free. The recording density had remained the same as that of the previous-generation 10,000rpm products, so the dual-platter model offered 73GB only. You just can’t have all at the same time. At each stage of technological progress the manufacturers have to balance between high spindle speed and high recording density (i.e. storage capacity). And users have to define their disk subsystem requirements clearly and realize that they have to compromise.

Again we have to admit that we don’t have the full selection of products in our today’s test session. Talking about Seagate’s models, we don’t have those of the three newest series. These are the 10K.3 (it has the same specs as the 300GB Hitachi), the 15K.2 (two times as large as its predecessor in terms of storage capacity), and the recently announced Seagate Constellation, which is a representative of the new breed of HDDs – 2.5-inch drives with a SAS interface but with a spindle rotation speed of 7200rpm.

Testbed and Methods

The following testing utilities were used:

Testbed configuration:

We formatted the drives 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. The drives were connected to an LSI SAS3041E-R controller. For all its highs, this controller has one obvious downside: all SAS drives are very slow when writing in FC-Test on it. We have to put up with this, though.

Performance in WinBench 99

We used the good old WinBench 99 for our low-level tests.

The next diagram compares the HDDs in terms of speed at the beginning and end of the full-capacity partitions created on them.

The three generations of 10,000rpm drives have ranked up in a logical way: each newer generation is about 50% faster than the previous one. What is more important, the newest generation with 150GB platters is about as fast as modern 3.5-inch drives with a spindle rotation speed of 7200rpm. Thus, we can expect them to be comparable when it comes to sequential reading/writing. The compact drives even look preferable as they have a smaller difference between the minimum and maximum speed: about 50% rather than 100% as usual. This is due to the smaller diameter of their platters.

The only 15,000rpm drive in this review has an interesting position. It is of course faster than the 10,000rpm drives with the same recording density but also faster than the models with two times its recording density. Thus, it is only inferior to the newer Hitachi C10K300. It is sad we don’t have a new-generation 15,000rpm drive for this review. It would be interesting to see if the newest 15,000rpm drives are still the fastest of all.

Performance in Intel IOMeter

Sequential Read & Write Patterns

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 be discussing graphs and diagrams.

It is all clear with the top speeds as we have seen the same standings in the previous test. It is more interesting to check out the performance of the HDDs on small data blocks before they reach their maximum. The Fujitsu is the first to accelerate to its full: it delivers its top speed on 2KB data blocks already, which is an excellent result. However, the new Hitachi delivers an even higher speed then, but it also has a higher top speed which it delivers from 4KB blocks onwards. By the way, the new Hitachi C10K300 looks very good against its predecessor. You can see a notable performance growth not only on small data blocks but throughout the entire test. When it comes to Seagate, the company’s 15K.1 and 10K.2 are comparable on small data blocks, being both superior to their predecessor. Unfortunately, the 15K.1 takes rather too long to reach its top speed, requiring 8KB blocks for that.

The writing results are roughly similar to those that we’ve seen at reading but the HDDs differ more on small data blocks. We can’t say that the Fujitsu is overall better then. The Seagate 15K.1 is a disappointment. Its speed is too low for its 15,000rpm till 8KB data blocks.

Disk Response Time

For 10 minutes IOMeter is sending a stream of requests to read and write 512-byte data blocks with a request queue of 1. The total of requests processed by the HDD is much larger than its cache, so we get a sustained response time that doesn’t depend on the HDD’s buffer size.

The Hitachi C10K147 and the Seagate 10K.2 turn in the best response time at reading among the 10,000rpm models but the gap from the others is quite negligible. As expected, the Seagate 15K.1 is unrivalled as its 50% higher spindle speed is hard to beat. It is interesting to compare the results of the 2.5-inch SAS drives with those of their 3.5-inch counterparts. Referring to the appropriate section of our previous roundup, you can see that the 2.5-inch products are but slightly inferior to their full-size opponents, the spindle rotation speed being the same. This is no wonder actually. It is only with the “slow” 2.5-inch drives with a spindle rotation speed of 5400rpm and 7200rpm that the access time is usually higher because their read/write heads are slowed down to reduce noise and the mass of the actuator. When it comes to server-oriented SAS drives, there is no talking about slowing anything down deliberately. Thus, the compact 2.5-inch models can work quite easily where a low response time is needed.

Most of them also deliver a good response time at writing. The Seagate 15K.1 is the only model to have some problems. It seems to have no deferred writing at all.

Random Read & Write Patterns

Now we’ll see the dependence between the drives’ performance in random read and write modes on the size of the data block.

We will discuss the results in two ways. For small-size data chunks we will draw graphs showing the dependence of the amount of operations per second on the data chunk size. For large chunks we will compare performance depending on data-transfer rate in megabytes per second. This approach helps us evaluate the disk subsystem’s performance in two typical scenarios: working with small data chunks is typical for databases. The amount of operations per second is more important than sheer speed then. Working with large data blocks is nearly the same as working with small files, and the traditional measurement of speed in megabytes per second becomes more relevant.

The speed of the disk subsystem with small data blocks depends only on the response time parameter, so the graphs lie close to each other, the Seagate 15K.1 with its 15,000rpm making the only notable difference.

When the data block size is increased to 128 kilobytes and larger, it is the sequential read speed that becomes important. Therefore the Hitachi C10K300 is successfully competing with the Seagate 15K.1 and even beats the latter on very large data blocks. The old Seagate 10K.1 looks poor against the others due to its low recording density.

Now let’s see what we have at writing.

It is the response time parameter that determines the drives’ results again but we can see a sudden performance hit of the Hitachi C10K147 and a smaller performance hit of the Fujitsu on 128KB data blocks.

The two mentioned products improve on larger data blocks, though. They take what places they deserve with their sequential write speed. The Hitachi C10K300 is first again as it has the highest-density platters. Take note that the high sequential speed (due to its high spindle speed) of the Seagate 15K.1 allows this drive to keep in second place despite high response and low recording density.

Database Patterns

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 the IOMeter: Database pattern.

We will build diagrams for request queue depths of 1, 16 and 256.

The HDDs go close to each other at the shortest queue depth. The Seagate 10K.1 is the only difference as it is always a little slower than the other 10,000rpm products. The Fujitsu is somewhat worse than its opponents at processing a large number of write requests.

The Seagate 15K.1 behaves in an interesting way. Having no rivals at reading, it is slower than the others at writing. This HDD just doesn’t seem to have any deferred writing at all.

When the queue is longer, the HDDs begin to exhibit their unique traits, i.e. their firmware algorithms. The Seagate 10K.2 and the Hitachi C10K300 are competing for first place among the 10,000rpm products. The former has a higher performance at high percentages of reads but the latter is better at writing. There are two drives that look worse than the others: the Seagate 10K.1 has modest deferred writing and the Fujitsu is too slow at reading.

The Seagate 15K.1 still does not want to show any trace of deferred writing. Having excellent results at reading, it sinks to fourth place at writing.

Hitachi’s HDDs, especially the new C10K300, are beyond competition at very long queue depths. Take note that they match the Seagate 15K.1 which has a higher spindle rotation speed. Of course, such high loads do not often occur in real servers, yet this is a good illustration of how firmware algorithms can affect performance.

Winding up this part of our tests, we will build a few diagrams that show the performance of each drive at five queue depths.

The Fujitsu MBB2 RC behaves like Fujitsu’s 3.5-inch SAS drives which produced similarly shaped graphs but with more effective deferred writing.

The two drives from Hitachi show a similar behavior but we can note that the newer C10K300 has a more aggressive and effective way of reordering requests: the graphs corresponding to medium and long queue depths are higher.

Interestingly, unlike the Fujitsu, Hitachi’s drives have a small but obvious performance growth at writing at a queue depth of 256 requests. The same is true for the lower queue depths.

The Seagate 10K.1 performs like the first 3.5-inch SAS drive from the same maker. It has low-efficiency writing and no performance growth at queue depths of 64 requests.

And like in full-size SAS drives, we see the next series, 10K.2, progress dramatically. Deferred writing is improved and the reordering of read requests is more effective. The processing of very long queues has not changed: the results are even lower at a queue depth of 256 requests than at 64 requests.

Funnily, but we have already seen Seagate drives behave like that. We are not 100% sure, but it looks like deferred writing is turned off while all requests undergo reordering (not only read requests as usual). This strategy was quickly abandoned in 3.5-inch drives, so the 15K.2 model is likely to behave in a more ordinary way.

Web-Server, File-Server and Workstation Patterns

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.

The Seagate 15K.1 is unrivalled under low read-only loads, which helps it get the highest overall score. Take note that the Hitachi drives are better than their opponents: the Seagate drives stop to accelerate at a queue depth of 64 requests, the Fujitsu does the same at 128 requests but the Hitachi drives keep on accelerating still.

When there is a small share of write requests in the load, the picture does not change much. The Seagate 15K.1 enjoys a smaller lead as it lacks deferred writing. The Seagate drives are also worse at high loads. The overall picture is the same as in the Web-Server pattern, though.

We don’t see big changes even at short queue depths and under variegated load typical of workstations. The Seagate 15K.1 is still better than the others but its advantage is now limited to the zone from 1 to 12 requests in the queue. We have a different loser now. This time it is the Fujitsu but not the Seagate 10K.1. Funnily, Fujitsu’s 3.5-inch SAS drives were among the best under this load.

When the test zone is limited to 32 gigabytes, the Hitachi C10K300 enjoys a huge advantage thanks to its denser platters. The Seagate 15K.1 cannot compete with the leader.

And once again we want to refer to our review of 3.5-inch SAS drives. By quickly comparing the performance ratings (which are actually necessary for the sake of comparison), you can see that 2.5-inch drives are not inferior to their full-size counterparts. Even with a smaller spindle rotation speed (in most models), they are quite competitive. It is only in the Workstation pattern with the reduced test zone that the 2.5-inchers are notably slower than the 3.5-inch drives.

Multi-Thread Read & Write Patterns

The multithreaded tests simulate a situation when there are one to four clients accessing the hard 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:

Ranking up according to the sequential speed when reading one thread, the HDDs do not change this order when processing multiple threads. The efficiency varies, though. The Hitachi C10K300 suffers more from the multiple threads than the other HDDs. The Seagate 10K.1 degenerates into random reading with very low performance at two and more threads. Alas, this behavior was typical for an entire generation of Seagate drives and we see it again now.

It is only with the single thread that the HDDs are ranked up like in the sequential write test. The Seagate 15K.1 is an exception, probably because it lacks deferred writing. The Seagate 10K.1 falls behind at two threads, too. This HDD seems to dislike multithreaded loads. A funny thing happens to the Hitachi C10K300. This drive survives the addition of a second thread easily but slows down heavily at three and four threads, falling behind its predecessor Hitachi C10K147.

Performance in FC-Test

For this test two 32GB partitions are created on the drive and formatted in NTFS and then in FAT32. A file-set is then created, read from the drive, copied within the same partition and copied into another partition. The time taken to perform these operations is measured and the speed of the drive 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 drive is working with two threads (one for reading and one for writing) when copying files.

You should be aware that the copying test not only indicates the speed of copying within the same HDD but is also indicative of the latter’s behavior under complex load. In fact, the HDD is processing two data threads then, one for reading and another for writing.

This test produces too much data, so we will only discuss the results achieved in NTFS. You can use the following link to view the FAT32 results.

Alas, the problems with the controller spoil the results. The Seagate 15K.1 and 10K.1 seem to win here but they have the lowest write speed according to the previous tests.

The Hitachi C10K300 is beyond competition. Its high recording density makes it unrivalled with every file-set. As for the others, the Hitachi C10K147 feels better than its opponents when processing the small files of the Windows and Programs patterns. The Seagate 15K.1 only enjoys an advantage over the 10,000rpm drives (excepting the 300GB Hitachi) in the ISO pattern.

When copying within the same partition or from one partition to another, the Hitachi C10K300 is in the lead, like in the reading test. The Hitachi C10K147 and the Seagate 10K.2 feel good, too. The Fujitsu is for some reason slower than the same-density HDDs when processing large files. The Seagate 15K.1 and 10K.1 are even worse, though. Their copying speed is no higher than 12MBps irrespective of the file-set we use.

Performance in PCMark 2004 / 2005

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 discuss 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 results can be learned from the table. The PCMark 2005 tests are:

The final result of the average of ten runs of each test.

Of course, these benchmarks, and the subsequent PCMark Vantage, are not so crucial for server disks as such HDDs are meant for different applications and loads. Anyway, let’s compare the HDDs in such tests, too. Click this link to view the table with results for PCMark 2004.

When copying files in PCMark, the HDDs deliver different results than in FC-Test. While the Hitachi C10K300 still has first place, the Fujitsu is second now. The pair of Seagate drives with low-density platters improves their results somewhat but cannot leave the last places.

The Fujitsu and Hitachi C10K300 cope better than the others with booting Windows XP up. It is firmware algorithms that determine a drive’s performance here: the Seagate 15K.1 with the lowest response time cannot show good results while the HDD with the highest-density platters is only second.

We’ve got the same standings in the Application Loading test. The Fujitsu is first, followed by the Hitachi C10K300.

The Fujitsu wins the General Usage test, too. The new Hitachi is now competing with the Seagate 10K2 for second place.

Scanning for viruses is highly sensitive to the caching mechanisms implemented in the drive’s firmware. This time, Hitachi’s HDDs cope best with it. Somewhat surprisingly, the Seagate 10K.2 and Fujitsu are on the losing side here.

There are no driver conflicts in this test and the HDDs are ranked according to their platter density. Thus, the Hitachi 10K300 takes first place, enjoying a 50% lead over the others. The Seagate 15K.1 is again hamstringed by the lack of deferred writing, delivering a very low speed.

Quite expectedly, the Hitachi drives take top places in terms of overall score, the new C10K300 enjoying a huge advantage. The Fujitsu is ahead of the Seagate HDDs although its performance in the individual tests was not consistent. The last place of the Seagate 15K.1 is a most remarkable thing. It means that a high spindle rotation speed is not so good if accompanied with low-density platters and poor firmware algorithms.

Performance in PCMark Vantage

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.

In the newer version of the benchmark and under multithreaded load we’ve got the same leader Hitachi C10K300 while the Fujitsu shows the strong points of its firmware and takes second place.

We see the same standings in the Gaming test. There is nothing to comment upon.

Photographers are going to prefer the Hitachi C10K300 to the other HDDs in this review because it is 50% faster than them. Moreover, the HDD even beats any of the 15,000rpm 3.5-inch drives we have tested so far.

Switching from Windows XP to Windows Vista does not change anything in the drives’ standings. The Fujitsu is still first, followed closely by the Hitachi C10K300. Take note that the Seagate 15K.1 is not too fast, again.

There are no changes in the Movie Maker test in comparison with the older version of the benchmark.

The Media Center test depends heavily on how the drive’s firmware interacts with the cache memory. Desktop HDDs usually show better results in it. Here, our SAS drives are not very fast. The old Seagate 15K.1 and 10K.2 are obviously worse than the others.

The HDDs split up into three groups under this load. The Hitachi C10K300 and Fujitsu are in the lead again while the Seagate 15K.1 is on the losing side. The latter seems to provide an example of how the older HDDs are worse than the newer ones.

There is one test where the 15,000rpm drive feels better than the others. It is the Application Loading test in this version of PCMark. The low performance of the Seagate 10K.1 is not a surprise but the Hitachi C10K147 finds itself in last place quite unexpectedly.

Anyway, the final scores have not changed much. The Hitachi C10K300 is still in the lead, followed by the Fujitsu.

Defragmentation

Next goes our homemade test of defragmentation speed. We created a very defragmented file system on a 32GB partition of a 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 disk we want to test. 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. We will run this test with AHCI enabled. For more information about it, you can refer to this article.

3.5-inch SAS drives are inferior to ordinary desktop HDDs in this test, yet we had not expected such poor results. The fastest drive here, the Fujitsu, takes almost 45 minutes to do the defragmentation. The two losers are especially impressive, though. The Hitachi C10K147 and Seagate 10K.1 spend two hours for this task! Take note that the spindle speed and recording density are not decisive factors in this test.

Performance in WinRAR

Now we are going to show 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 drive. This test depends heavily on CPU performance, but the storage device affects its speed, too.

The HDDs are all good in the archiving test. The difference between the best and worst model is a mere 21 seconds. The Fujitsu and the Hitachi C10K300 look better than the others, again.

The results are not so uniform when the HDDs unpack the archive. There are two obvious losers: the Seagate 10K.1 and the Hitachi C10K147. Surprisingly, the Seagate 15K.1 takes first place. At the end of our test program there is one more test where that model can win.

Power Consumption

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.

When it comes to starting up, the HDDs have a rather standard power draw on the 5V line. As opposed to that, they consume a very low current from the 12V line. Unlike HDDs for notebooks, high-performance SAS products cannot do without 12V, though.

The Seagate 10K.1 is poor in comparison with the others. It needs a 50% higher current than what the other HDDs require. Even the 15,000rpm model needs less. The Hitachi C10K300 and the Fujitsu look best in this test, just like in many previous ones.

We’ve got one leader under low load. It is the Fujitsu and it has the most economical mechanics. Second place goes to the Seagate 10K.2 which has the most economical electronics. The Hitachi drives are not very good here. They have very voracious mechanics and their electronics is not exactly economical, either.

These compact HDDs look excellent in comparison with 3.5-inch counterparts and consume only half as much power.

The small diameter of the platters and the reduced heads (but rather fast heads nonetheless) produce considerable benefits when it comes to reducing power consumption. The random reading of the best products, the Seagate 10K.2 and the Hitachi C10K300, fits within 6.5 watts. The Seagate 15K.1 with 15,000rpm platters fits within 8 watts while its full-size opponents need up to two times as much as it.

Take note that the consumption of the electronics contributes a lot to the overall result of compact SAS drives in this test. The Seagate 10K.2 consumes more from the 5V line than from the 12V one. The other models consume about the same amount of power from both lines.

The overall picture remains the same when we switch from random reading to random writing. The HDDs all consume the same amount of power from both lines, and the newer models are superior to the older ones in terms of power efficiency.

It is more interesting at sequential reading. The read/write heads do not have to move so actively and the 12V consumption is lower while the electronics works heavily at sequential operations. We can note that each new generation of HDDs have both more economical electronics and mechanics. The overall standings are the same: the Hitachi C10K300 and the Seagate 10K2 are in the lead while the old Hitachi C10K147 and Seagate 10K.1 are the most voracious drives in this test.

Take note that the Seagate 15K.1 is quite comparable to the 10,000rpm drives in terms of power consumption.

When we switch from sequential reading to sequential writing, the overall picture still remains the same: newer HDDs need less power and the drives mostly consume from the 5V line, i.e. with their electronics.

Conclusion

We did not compare HDDs of different form-factor in this review directly, but the performance of the 2.5-inch models seems to be quite comparable to that of their full-size counterparts. Moreover, the 2.5-inchers are faster than 3.5-inch drives at random-address operations due to the smaller working zone of the platter.

Talking about the specific models, we want to single out the Seagate Savvio 15K.1 for its excellent performance in server tests. The higher spindle rotation speed (15,000rpm) gives it an edge against the other HDDs in tests that focus on input/output operations.

The Hitachi Ultrastar C10K300 should also be praised for its high recording density and excellent performance in practical applications. It won all the sequential-load tests and FC-Test and delivered good results in the server patterns and PCMark.

Unfortunately, we did not have an HBA controller with support for SAS 6Gbps at the time of our tests, so the key feature of the Hitachi Ultrastar C10K300 – its support for the new version of the SAS interface – is left unexamined. But we hope it is not the last of our reviews of 2.5-inch SAS drives and we will be able to return to this issue soon.