by Aleksey Meyev Nikita Nikolaichev
10/13/2009 | 11:08 AM
Soon after our recent review of four solid state drives from Intel and Corsair called “Hi-Speed Chase: Corsair P128 vs. Three Intel SSDs”, we got as many as six new models. To be specific, it is a second-generation Intel X25-M model based on 34nm chips (the first generation was based on 50nm technology) and five SSDs from OCZ’s Agility, Summit, Vertex, Vertex Mac Edition and Vertex Turbo series. Let’s check them out right now!
First goes the first-generation (G1) Intel X25-M which proved some time ago that SSDs had grown from ugly ducklings into swans already. It will serve us as a reference point in this test session.
The first generation of Intel’s SSDs is leaving already, though. It is being replaced with a second generation that is based on 34nm rather than 50nm flash memory chips. The controller has been revised, too. We should not expect any performance breakthroughs from the transition, though, and the specified speeds have not actually changed: 250MBps for reading and 70MBps for writing. It is storage capacity that is going to grow up. Early X25-M products used to have a capacity of 80GB and then grew to 160GB whereas the second generation begins with a 160GB model, a 320GB version coming out in near future. The exterior of the SSD is interesting, too. The first-generation 80GB drive had a full thickness of 9.5 millimeters while the second-generation SSD and the first-generation 160GB model are somewhat slimmer. There is a removable plastic frame that makes their metallic case as thick as standard.
We do expect some performance benefits, though. Why? First off, the capacity of the controller’s RAM chip is doubled (from 16 to 32MB). And the new generation has got new firmware. Unfortunately, the release of new products was somewhat spoiled by the fact that the firmware of the first batch had a bug that might cause a loss of data. Intel reacted instantly, though. The corrected firmware was quickly made available (it also suits Intel’s first-generation SSDs, by the way). And of course, all new batches of second-generation SSDs that come to retail have the updated firmware. As far as we know, they also come in different packaging. Instead of the modest cardboard box you can see in the photo above, the new packaging resembles a CPU box. Besides a disc, the kit will also include an adapter to install the SSD into a 3.5-inch system case bay.
OCZ has taken to SSD production most seriously. Right now, the company offers as many as nine series of SSDs in the form-factor of 2.5-inch drives. Besides, it supplies models designed as expansion cards with PCI Express and miniPCI Express interfaces. We will start from the most popular series called Vertex.
Like all other SSDs of this firm, the drive comes in a neat box like a jewel, kind of hinting at the product’s price. OCZ’s Vertex series drives are based on the Indilinx Barefoot controller we have not yet tested in our labs and on Samsung flash memory chips. They are also equipped with 64 megabytes of cache clocked at 166MHz. The manufacturer claims the SSD can deliver peak read and write speeds of 270MBps and 210MBps, respectively. The sustained write speed is declared to be 120MBps, though. These numbers refer to the 250GB model whereas the other models have lower speeds of both reading and writing. The junior models probably have fewer controller channels and a smaller pool of free memory cells. The smaller this pool, the harder it is for the controller to clean up space to write new data.
An interesting feature of all SSDs based on Indilinx controllers is that they are supported by the Indilinx Wiper Tool. This program can be downloaded as OCZ_GC.exe from the OCZ website. It is a simple and effective tool. Here is what it does: today’s operating systems cannot tell an SSD that a file is removed and what LBA addresses are freed at that. They only write changes into the file allocation table. This is all right with HDDs but the performance of SSDs can degenerate if a previously loaded SSD is being written to. The mentioned tool asks the OS about the status of each LBA address pertaining to the SSD and tells the latter what LBA addresses are free from the file system’s point of view. As a result, the SSD can erase data from appropriate memory cells beforehand rather than just prior to writing into them. Of course, that’s nothing but a crutch that is needed to solve the problem of interaction between the OS and the new type of storage devices. But before Windows 7, which will do the same thing on the OS level, this tool is better than nothing.
The OCZ Vertex has firmware version 1.3. Version 1.4 is beta as yet.
Frankly speaking, we don’t know about any differences of this SSD from the previous model. As far as we know, Apple does not put forth some special requirements to SSDs. The device’s case contains the same stuff as the ordinary Vertex series drive: an Indilinx Barefoot controller, Samsung flash memory chips, 64 megabytes of cache memory clocked at 166MHz. The specified speeds are somewhat lower: 240MBps for reading and 150MBps for writing. The speed of sustained writing is 100MBps. There must be some changes in the firmware whose version is 1.31.
Here is one more OCZ drive with the word Vertex in its name and based on the Indilinx controller. The distinguishing features of the Vertex Turbo are easy to see, though. Its memory and controller work at higher frequencies. This must be the reason for this controller to have only one firmware version, 1.0. We will see if the minor increase in frequencies (e.g. the memory frequency is overclocked from 166 to 180MHz) can affect the SSD’s performance or the multiple changes in the firmware of the ordinary OCZ Vertex have a bigger effect. The specified speeds are the same 270-210-120MBps at least.
The Agility is one more SSD from OCZ based on an Indilinx controller. OCZ refers to all Vertex series drives (irrespective of any suffix in the model name) as to performance products whereas the Agility series is referred to as mainstream. This SSD probably uses cheaper memory and fewer memory channels. The maximum storage capacity of the Agility series is 120GB and the specified speeds are lower than those of the same-capacity Vertex: 230-135-80GB. The firmware is version 1.3.
The OCZ Summit drive comes from a performance series but is based on Samsung’s latest platform rather than on Indilinx’ controller. We know the platform by the Corsair P128 SSD as consisting of a Samsung S3C29RBB01 (PB22-J) chip together with 128 megabytes of cache. The specified read and write (both peak and sustained) speeds are 220MBps and 200MBps. The firmware version is VBM1801Q.
We use the following method of benchmarking SSDs:
We guess this method will help compare the SSDs under identical conditions and yet produce real-life results.
The following testing utilities were used:
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 the mainboard’s SATA port in each test and worked in AHCI mode.
We will use WinBench 99 for low-level tests. This benchmark goes first and we “organize” the flash memory cells before it (by a stream of read requests as the SSD manufacturers suggest), so it is indicative of the maximum speed you can have with your SSD. Let's take a look at the read graphs first:
It must be noted that the graph of the first-generation X25-M does not show the peak speed but rather the speed of the “unleveled” drive. This is indicated by its shape. The other SSDs show a rather consistent picture: the Vertex Turbo is first, followed by the two other Vertex series SSDs and the new Intel which deliver the same results. The Summit based on the Samsung controller is somewhat slower, and the Agility is at the bottom of the diagram (quite expectedly as it is actually positioned as a simplified and inexpensive SSD).
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 processed 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 SSDs split into two groups when processing small data blocks: both SSDs from Intel and the Samsung-based OCZ Summit are somewhat faster than the Indilinx-based products. The new Intel and the Vertex Turbo deliver the highest top speed, being ahead of the ordinary Vertex and Mac Edition. The Summit is somewhat slower but better than the first-generation Intel. Take note of the obvious progress Intel’s SSDs have made in processing large data blocks. The Agility is the slowest drive in this test, yet it wouldn’t be inferior even to the best of modern HDDs.
We can see two groups of drives at writing, too. The faster group includes the Vertex Turbo, Vertex Mac Edition and Summit. Interestingly, the ordinary Vertex falls behind its series mates, being even somewhat slower than Intel’s SSDs. The latter deliver their specified speeds, the second-generation model being a tiny bit better than its predecessor. The Agility is nothing but modest in this test. Although it is not promised to be fast, it is based on the same Indilinx controller as the Vertex series. Frankly speaking, we had expected it to be better.
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 amount of requests processed by the drive is much more than its cache buffer, so we get a sustained response time that doesn’t depend on the drive’s cache.
Intel’s SSDs of both generations are unrivalled in terms of response time. They are very fast both at reading and writing. All of the Indilinx-based drives are good, too. They all have the same read response time. And when it comes to writing, it is only the simplified Agility that has a worse, but not downright poor, result than the others. The Summit, even though the worst model in this comparison, is not actually bad. Its read response is 100 times as good as that of 7200rpm HDDs! And its write response is five times as good as HDDs’. So, we already know at least three controllers that allow SSDs based on slower MLC flash memory to deliver a better write response time than HDDs have. Early SSDs could not do that.
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.
Let’s start with reading.
IOMeter: Random Read, operations per second
Intel’s SSDs are unrivalled at reading small random-address data blocks. OCZ’s Indilinx-based drives are only half as fast (and the Agility doesn’t differ from the others in this respect). The transition to the newer tech process has made Intel’s product somewhat better in this test than before. The OCZ Summit is the last one again: it is slower than the other OCZ drives on smallest data blocks. It looks like its controller cannot cope with processing such a huge number of very small requests.
IOMeter: Random Read, megabytes per second
The Summit joins the other SSDs from OCZ when processing large data blocks whereas the Agility falls out of that group. The Agility really seems to have fewer controller channels. As the data block size is getting larger, the OCZ drives catch up and even outperform the first-generation X25-M but Intel has left no chance to its opponents by releasing the updated model whose performance is unrivalled.
IOMeter: Random Write, operations per second
We’ve got a very interesting picture here. Intel’s drives are obviously better than their opponents but the first-generation model is faster at processing smallest data blocks. This must be due to changes in firmware. Hopefully, the difference is not a bug but the result of optimizations of the drive’s performance in other modes. Anyway, those 10 thousand operations per second is very, very good for an MLC-based solid state drive. The SSDs with other controllers cannot yet match this result. OCZ’s Indilinx-based products perform differently in this test as they have different firmware. The Vertex Mac Edition is the best of them, perhaps because its firmware version is higher. The OCZ Agility is slower in comparison with the others. It is a low-end product indeed. However, it is still faster than HDDs whereas the OCZ Summit is downright poor. If you compare its results with those of the Corsair P128, you can see that the Corsair is much better even though these SSDs share the same platform.
IOMeter: Random Write, megabytes per second
Both models from Intel hit against their maximum of 70-80MBps and give way to the Vertex Mac Edition which delivers a top write speed of 200MBps. Funnily, this is the only Indilinx-based drive that confidently copes with writing large random-address data blocks. The performance of the ordinary Vertex and the Turbo fluctuates wildly, obviously due to imperfect firmware. The Agility is stable and predictable but too slow.
The OCZ Summit is a disappointment again. Its writing performance is too low.
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.
We’ve got two leaders at the minimum queue depth. The second-generation X25-M is a considerable improvement over its predecessor and is absolutely unrivalled at high percentages of reads. The Vertex Mac Edition is ahead of the others at writing. There must be some fundamental changes in its firmware or some other reason, but the other Indilinx-based models behave differently in this test and deliver much lower performance. Interestingly, the Agility hardly differs from its opponents under such loads.
The Summit is a disappointment once again. It is competing with the other OCZ drives at high percentages of reads but slows down awfully when there is some share of writes to be performed.
When the queue depth is increased to 16 requests, Intel’s second-generation drive goes even further ahead at high percentages of reads. It is two times as fast as its closest pursuer! Just think about it: this SSD can process 25 thousand requests to read 8KB data blocks in a single second! Before the arrival of solid state drives such performance could only be achieved by creating virtual disks in system RAM. Even the iRAM device, based on DDR SDRAM, was only half as fast.
There is nothing particularly interesting at the longer queue depth. We still have two leaders at different types of load, the others pursuing them more or less successfully.
Winding up this section of the review we will build diagrams showing graphs for five different queue depths for each SSD.
Intel’s engineers have done a good job writing firmware for their second-generation SSDs. The increased efficiency of NCQ algorithms is obvious. The newer model is obviously faster. We have doubts about whether this is true NCQ, though. The SSD can hardly reorder requests according to their LBA addresses because this would produce no performance benefits. We guess that it reorders requests in such a way as to load all the controller channels as much as possible simultaneously. It is also good that the SSD does not slow down too much at very high percentages of writes.
The OCZ Summit’s diagram resembles the one of the Corsair P128. This platform seems to have a high degree of repeatability. We can see that the SSD gets faster at reading when the queue depth is increased, but suffers a terrible performance hit at writing. There is one difference from the Corsair P128, though. The OCZ Summit successfully copes with the queue depth of 4 requests. It looks like the SSD runs out of buffer memory for write requests at long queue depths.
OCZ’s Agility, Vertex and Vertex Turbo behave in a similar manner. They enjoy a small performance growth when the request queue is increased, but only at reading. The Vertex Turbo differs from the others with its repeatable and inexplicable performance jump at about 70% writes.
The OCZ Vertex Mac Edition differs from the other Indilinx-based products, making us suspect that it uses some secret controller. Its performance increases at longer queue depths even at high percentages of writes. More importantly, this SSD is generally better than its mates at writing. Our praise goes to the developers of its firmware if this is due to the firmware. It is a fantastic result for a drive based on MLC memory to deliver over 4 thousand operations per second under any load!
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.
After the Database tests it is quite easy to predict the winner of this pattern which consists of read requests only. The second-generation X25-M enjoys a huge advantage over its opponents. Moreover, it differs from its predecessor in not having a performance hit at long queue depths (all thanks to the new firmware!). It is simple with the other drives: the Vertex Turbo has somewhat better results than the other Indilinx-based models while the Agility is the slowest among them. The OCZ Summit is the worst drive in this test.
The standings change when there appear write requests, but the leader remains the same. Intel’s new generation of solid state drives is a perfect choice for servers! The best of the opponents, the OCZ Mac Edition, can only rival Intel’s first-generation products. The other Indilinx-based models have lower performance, and the Summit is downright poor at writing. On the other hand, single hard disk drives rarely score over 100 points in this test while the Summit scores 493. A RAID array built out of eight SAS disks could deliver over 1000 points, but the X25-M G2 has over 12 thousand points!
It’s like in the previous test: the new Intel is unrivalled and, unlike its predecessor, has no performance slumps. The OCZ Vertex Mac Edition stands out among Intel’s opponents. Two things must be noted here. First, the Vertex Turbo is the worst drive with Indilinx controller (probably because it has the oldest firmware). And second, the Summit behaves most characteristically. Its performance does not grow up along with the request queue as that of Intel’s SSDs and does not remain on the same level as that of the Indilinx-based models, but declines.
When the test zone is limited to 32 gigabytes, the SSDs all slow down somewhat but their standings remain the same.
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:
Theoretically, flash-memory drives should pass the multithreaded reading test with minimum performance loss because they don’t have to spend time moving any read/write heads around like HDDs do. However, there is a performance hit in practice. The second-generation X25-M is the only drive that has no problems with multiple read threads. Its speed even grows up somewhat. Its predecessor copes with the test well enough, too. OCZ’s products are much worse: they only deliver two thirds of their normal speed when reading two and three threads. The Summit, based on a Samsung controller, goes on slowing down at four threads, too, while the Indilinx-based models speed up then: two of them return to their normal speeds and the Vertex Turbo and Agility even set personal records. Perhaps this has something to do with four-channel controller architecture, but the reduction of speed at two and three threads is inexplicable.
Most of the SSDs are not good at multithreaded writing which requires efficient caching. The X25-M are the only drives to deliver stable speed. Being fast at processing one thread, the products from OCZ get slow at two and more threads. As a result, they slow down to the level of Intel’s drives or even lower.
For this test two 32GB partitions are created on the SSD and formatted in NTFS and then in FAT32. A file-set is then created, read from the SSD, copied within the same partition and copied into another partition. The time taken to perform these operations is measured and the speed of the SSD 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 SSD 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 Indilinx-based SSDs are ahead when writing large ISO files. They don’t deliver the promised speeds but all of the Vertex series models offer 100MBps and more, the Agility following closely behind. They slow down on small files, especially those of the Program pattern, and Intel’s SSDs take first two places. By the way, Intel’s second-generation model is always 10MBps better than its predecessor. The OCZ Summit is the main disappointment of this test as its performance is low with each file-set.
All of the SSDs are excellent at reading files, especially the Programs file-set where they are about two times as fast as modern HDDs. The new Intel X25-M is the overall winner while the Summit and Agility are losers.
Intel’s SSDs are ahead at copying, the newer model being some 10MBps better than the older one again. The Summit stands out among the others. It is surprisingly good with large ISO files but does not like the Install pattern. The Vertex Turbo is worse than the others, probably due to imperfect firmware.
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.
PCMark 2004 results table
PCMark04 says that the OCZ Summit is faster at copying files than Intel’s SSDs. Otherwise, the standings are the same: the Indilinx-based models go close to each other, the Vertex Turbo being somewhat slower than the others.
The X25-M is better than the others at booting Windows XP up. Interestingly, its predecessor is the worst under this load. The Summit stands out among OCZ’s products.
The OCZ Summit is the best in the Application Loading test, outperforming the second-generation X25-M and OCZ Vertex that are contending for second place. The Vertex Turbo shows the worst result: the increased controller frequency cannot make up for imperfect firmware.
The Summit keeps its first place in the General Usage test whereas the Vertex Turbo is still the slowest of all. The new X25-M is much better than its predecessor, but still cannot beat the Indilinx-based models with newer firmware.
Five out of the seven SSDs pass the Virus Scan test excellently. The OCZ Summit and the older X25-M are slower than the others here.
The File Write trace seems to be based on large files: the Indilinx-based models are in the lead, delivering high speeds, whereas the two SSDs from Intel are on the losing side. The second-generation model from Intel is somewhat faster again.
The OCZ Summit has the highest overall score. The Indilinx-based drives, except for the Vertex Turbo, are all better than the second-generation X25-M in terms of overall performance rating. The older X25-M has the lowest score.
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.
This multithreaded load test agrees with IOMeter: Intel’s SSDs are better than their opponents when reading multiple data threads.
Intel’s SSDs are also better than their opponents under gaming load.
Curiously, the simplified OCZ Agility is about as fast as its full-featured Indilinx-based counterparts. It looks like its “simplification” won’t be conspicuous in real applications.
Intel’s SSDs cope with the photo gallery better than the others, too. Interestingly, the Vertex Turbo differs from the others either due to its increased frequencies or to the older firmware being surprisingly good under this particular load.
Quite unexpectedly, the first-generation X25-M takes first place when booting Windows Vista up. Its newer mate is only third, behind the Vertex Mac Edition. The Vertex Turbo performs poorly here, falling behind the simplified OCZ Agility even.
There are three leaders in this Movie Maker test: the Vertex, Vertex Mac Edition and Agility. They owe this success to their firmware because the Vertex Turbo performs poorly although is based on the same controller. It is the OCZ Summit, based on the Samsung controller, that has the lowest result in this test, though.
This test is traditionally sensitive to caching. Here, the Indilinx-based SSDs with new firmware, namely the Agility, Vertex Mac Edition and Vertex, take three top places with similar results. Like in the previous test, the OCZ Summit is the slowest drive of all.
Intel’s products are superior here. The OCZ Summit and the Vertex Turbo are losers.
We have the same winners when loading applications and Intel’s SSDs are far faster than their opponents here. The Vertex Turbo suffers from its older firmware again.
This version of the benchmark gives two top places to Intel’s products, the second-generation model having a somewhat higher score. Interestingly, the simplified Agility is but slightly slower than the Vertex and Vertex Mac Edition whereas the Vertex Turbo shares last place with the Summit.
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. The tested disk 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.
Strictly speaking, this test makes no practical sense for solid state drives because there is nothing to defragment on them. Every memory cell is equivalent to any other, so defragmentation won’t have any effect. However, this test will allow us to compare how much time SSDs spend reading and writing the same amount of small data blocks.
Intel’s SSDs are superior in this test. The newer X25-M takes only half the time the best of the OCZ models needs to complete the test. The first-generation drive from Intel is almost as fast as the leader, too. The OCZ Summit, based on a Samsung controller, is third. The Vertex Mac Edition and Agility are the best among the Indilinx-based models whereas the Vertex Turbo and the ordinary Vertex take over half and hour to pass the test. This is a vivid example of how performance depends on firmware.
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.
Intel’s SSDs are better than the others at reading numerous small files and then writing one large file. The models from different generations notch almost the same time here. The Indilinx-based products from OCZ vary greatly: the Agility is ahead of all OCZ SSDs, including the Summit with Samsung controller, while the Vertex and Vertex Mac Edition are slower than it by almost one minute.
When the SSDs unpack the archive, the standings are not neat, either. For example, the Vertex Mac Edition is in the top three, sharing the podium with Intel’s products. The other Indilinx-based models took longer to perform the task. The OCZ Vertex is especially poor: 2 minutes is too long a time for such a simple task.
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 OCZ Summit differs from the others in terms of start-up current. The Samsung platform seems to have a very modest appetite. Interestingly, the updated Intel X25-M is in last place whereas its predecessor is second: a transition to a thinner tech process is usually associated with a reduction in power consumption.
The SSDs almost do not consume anything in idle mode at all. Intel’s SSDs are the most voracious in this mode but their power draw is less than 1 watt. The OCZ Summit needs even less than one third of a watt!
Intel’s drives require more power than the others at random loads, too. Their advanced controller needs quite a lot of power (in comparison with other SSDs, of course). We can note, however, that the second-generation X25-M has the same consumption at reading as the first-generation one but is much more economical at writing and doesn’t differ much from its opponents then. Take note that while HDDs need more power to do reading, SSDs need more power to do writing. It’s because flash memory has to do more operations and the controller is always busy.
There are two leaders at sequential operations: the OCZ Summit and the Vertex Mac Edition. It is hard to tell why but the other models with the same controller need more power at writing. The Vertex Turbo is the only SSD that needs more power at reading than at writing. We can only explain this by the increased frequencies of this SSD’s processor and memory.
As for Intel’s products, the second-generation SSD has a somewhat lower power draw.
Overall, SSDs are very economical and consume somewhat less power than 2.5-inch HDDs. They are obviously unrivalled in terms of efficiency per watt.
Frankly speaking, it is not easy to name the winner of this test session, but the second-generation Intel X25-M deserves this title more than anyone else. Intel has endowed its SSDs with new 34nm chips with considerably revised firmware. Being somewhat inferior to its predecessor at writing very small random-address data blocks, the new X25-M delivers excellent and very stable results under server loads. It is also good under most loads typical of home computers, which makes it a universal and very fast storage device. Still, we’d prefer to call it server-oriented just because it has no rivals in server applications.
However, it no time for resting on laurels as Intel’s products do meet some serious competition from SSDs based on the Indilinx controller. Their performance under most loads is but slightly lower than that of the X25-M but they have a much higher speed of linear writing. The performance of Indilinx-based models varies significantly depending on firmware, remember the excellent results of the OCZ Vertex Mac Edition in IOMeter: Database and at random writing. Therefore it is sad that the Vertex Turbo does not have new firmware. With the basic firmware this highly perspective SSD is not very interesting, unlike the OCZ Vertex and OCZ Vertex Mac Edition. The latter is quite a surprise to us, by the way. We didn’t expect the Mac version to be so different from the others.
People who do not need large storage capacities may be interested in the OCZ Agility. It is cheaper than the other SSDs while its performance is decent due to the full-featured Indilinx controller inside.
The OCZ Summit, based on Samsung’s platform, is expectedly similar to the Corsair P128 we tested earlier. It is the most power-efficient SSD and works well with large files, but loses to its opponents under server loads and at writing small data blocks. Still, it uses the latest generation controller, which makes it much faster than today’s HDDs.
Thus, we have already tested SSDs with three popular controllers in our labs. We only lack an SSD with JMicron’s controller for the picture to be complete.