When Windows Vista arrived more than three years ago, Microsoft’s then-flagship desktop operating system brought with it a number of new technologies designed to improve performance. SuperFetch promised to deliver performance gains by caching frequently access data in unused portions of main memory. ReadyBoost took a similar approach, but rather than filling free memory opportunistically, it tapped the unused capacity of USB flash drives. Then there was ReadyDrive, which offered support for hybrid hard drives that combined flash memory with traditional mechanical platters.
In theory, hybrid hard drives make perfect sense. Flash memory offers essentially instantaneous seek times that are at least an order of magnitude quicker than the fastest mechanical hard drives, but it’s also quite expensive. That’s why SSDs have relatively low capacities and rather unattractive cost-per-gigabyte ratios. Traditional hard drive platters can offer substantially more storage capacity at a much lower price, but their inherent mechanical latency makes for much lengthier access times. Meld those technologies together, and you should be able to enjoy the best of both worlds.
I had high hopes for hybrid hard drives after Vista’s launch. However, that was to be the last I heard about ReadyDrive. Samsung and Seagate did end up bringing hybrid drives to market, but neither company’s offerings gained much traction. In retrospect, I’m not surprised. Being hitched to the most reviled operating system since Windows ME surely didn’t help ReadyDrive’s chances. With most users and businesses happy to stick with Windows XP, the potential market for hybrid drives may have ended up being much smaller than initially hoped.
The hybrid storage concept didn’t die out as ReadyDrive faded into obscurity. As the price of solid-state drives declined, PC enthusiasts began rolling their own hybrid storage setups by combining low-capacity SSDs with mechanical hard drives. Such configs really do deliver the best of both worlds. The SSD can be used to house OS and application files that need to be accessed quickly, while the mechanical drive provides ample capacity for data that just needs to be stored.
The do-it-yourself hybrid approach might work well for desktops, but few notebooks can accommodate the two drives required. Fortunatelyand wisely considering the market’s shift toward notebooksSeagate didn’t abandon the hybrid concept entirely when it gave up on ReadyDrive. Instead, the company set out to devise an OS-independent hybrid technology that appears in the new Momentus XT, which combines 500GB of 7,200-RPM mechanical storage with 4GB of flash memory. Naturally, we had to sample this hybrid redux for ourselves.
Before exploring the new hybrid direction Seagate is taking with the Momentus XT, it’s worth delving into the concept’s ReadyDrive roots. As one might expect, ReadyDrive was designed to populate a hybrid drive’s flash memory with frequently accessed data. Data associated with system startup and hibernation are also loaded into the flash because, according to Microsoft, the system often waits on traditional hard drives during those tasks.
There’s more to ReadyDrive than speculative caching to improve performance, though. Microsoft wanted hybrid drives to offer power savings, so ReadyDrive can spin down a drive’s platters and operate exclusively out of its flash memory. If a read request can be serviced by the flash’s contents, there’s no need to wake up the platters. Writes can also be cached in the flash in this low-power state, with the platters springing into action only if the flash doesn’t have enough free capacity to accommodate a write request.
When Microsoft architect Ruston Panabaker presented ReadyDrive at WinHEC in 2006, he called for a minimum flash capacity of just 50MB. 120-256MB was recommended as the minimum necessary to realize “significant” performance and power consumption benefits. That might not sound like much, but keep in mind that Vista was in development at a time when flash memory was considerably more expensive than it is today. During that era, we reviewed an IDE solid-state drive that cost a whopping $320 for just 8GB. Fast-forward to the present day, and $320 will get you a Kingston SSDNow V+ SSD with 16 times the capacity.
According to Seagate, ReadyDrive tried too much to be all things to all people. The hard drive maker wasn’t crazy about being tied to a single operating system, either. Even if Vista hadn’t been a pariah, hybrid hard drives are particularly appealing in the mobile world, which just happens to be filled with a whole lot of MacBooks running OS X. Besides, Seagate contends that a hard drive can make better decisions about what to put in its flash because the drive knows more than the operating system about how data is arranged on the disk.
Into Adaptive Memory
Seagate refers to the intelligence that populates the Momentus XT’s flash as Adaptive Memory, cleverly disguising what is ultimately a massive read cache. Like ReadyDrive, Adaptive Memory populates the flash with data from the disk. However, Adaptive Memory is the only mechanism that writes to the flash. Writes from the host system never touch the flash memory, instead proceeding to the disk through the drive’s DRAM cache, just like in a traditional hard drive.
Rather than looking at user access patterns at the file level, Adaptive Memory tracks Logical Block Address (LBA) access patterns to determine what gets cached. Access frequency is considered by the algorithm, but surprisingly, it’s not the most important factor. Seagate maintains that not all data can be accessed significantly faster from the flash than it can be pulled from the platters. Flash may be substantially quicker than mechanical media with small random reads, but the gap between the two is much tighter with larger sequential transfers, especially if you don’t have a lot of flash chips to spread across a multi-channel array. That’s why Adaptive Memory puts only the data it thinks will deliver the biggest performance gains into the flash. When the flash is filled to capacity, access frequency becomes more of a determining factor.
Naturally, it takes some time for Adaptive Memory to learn a user’s tendencies. Seagate says that a task must be performed two to three times before Adaptive Memory has the flash’s contents optimized for it. Adaptive Memory is a relatively quick study, then, but also one that’s prone to memory loss. Because it reorganizes data on the disk, defragmenting the Momentus XT will reset everything that Adaptive Memory has learned.
Adaptive Memory caches data in the Momentus XT’s flash in chunks as small as 2KB. That granularity nicely matches the flash’s custom formatting, which uses a 2KB page size instead of the 4KB pages found inside most SSDs. Each flash block still has 128 pages, but because they’re smaller, the block size is only 256KB as opposed to 512KB. This smaller block size should lessen the impact of the block-rewrite penalty by reducing the size of a single block rewrite. That said, the block-rewrite penalty is really a non-issue for the XT because the host system is never waiting on a flash write. The block-rewrite penalty will only affect how quickly Adaptive Memory can transfer data between the platters and the flash.
Unlike most solid-state drives, which employ a multitude of multi-level cell (MLC) flash modules, the Momentus XT makes do with a solitary single-level cell (SLC) chip that weighs in at 4GB. SLC memory is common in enterprise-class SSDs, where its order-of-magnitude advantage in write-erase endurance over MLC flash offers greater longevity. Obviously, Seagate doesn’t want Adaptive Memory to burn too quickly through the 10,000 write-erase cycles that typical MLC chips can handle.
SLC flash also tends to have faster write speeds than MLC. Seagate says that the XT’s flash has peak reads speeds of 140MB/s and peak write speeds of 100MB/s. That’s pretty spry for a single chip.
The XT’s flash memory is joined by a 32MB DRAM chip that offers twice the capacity of the cache used in Seagate’s previous 7,200-RPM mobile flagship, the Momentus 7200.4. The capacity bump is supposed to speed write performance, although it’s likely a way for Seagate to differentiate the XT from more mainstream Momentus drives, as well. On the desktop, the Barracuda XT has twice the cache of other ‘cudas.
A single Seagate drive controller governs the XT’s cache, flash, mechanical platters, and host interface. Seagate elected to stick with a 3Gbps Serial ATA interface, which makes perfect sense given the drive’s mobile ambitions. Besides, the XT isn’t fast enough to require more than a 300MB/s SATA link.
On the mechanical front, the XT features a pair of 250GB platters spinning at 7,200 RPM. That’s the same platter configuration as the Momentus 7200.4 we reviewed a year ago, so it’s not exactly cutting-edge technology. Indeed, Seagate has already announced a new 750GB mechanical Momentus model that packs two 375GB platters with the same 7,200-RPM spindle speed as the XT. Seagate says a different design team was responsible for the XT, which is why that drive isn’t using the new platters.
With a much higher areal density, Seagate’s 375GB platters should be capable of sustaining higher read and write speeds than the 250GB disks inside the Momentus XT. We were supposed to receive one of the new Momentus 750GB drives with our XT sample, but we’ve now been told that the high-capacity drive won’t arrive until mid-June. Perhaps Seagate wanted to a avoid having its new mobile flagship compared with a drive that will almost certainly offer superior sequential throughput.
The XT’s platters are claimed to be capable of hitting speeds of up to 153MB/s, nicely illustrating Seagate’s rationale for not caching everything in the flash, which reads at 140MB/s. Seeks to the flash will be nearly instantaneous, while it’ll take an average of 12 ms to seek data off the disk. A dozen milliseconds may sounds nearly instantaneous, too, but keep in mind that hard drives live in a world where multiple processor cores tick away at billions of cycles per second.
You’ll notice that I haven’t spent any time talking about Adaptive Memory’s advanced power-saving features. That’s because Seagate is using the flash primarily to improve performance rather than to reduce power consumption. The drive will spin down the platters to a standby mode after a minute of inactivity. If a read request can be completely serviced by the contents of the flash memory while the drive’s in standby mode, there’s no need to spin up the platters. However, a missed read or any write operation will kick the platters into gear again. Adaptive Memory won’t cache data based on whether having it in the flash is likely to extend the time the platters spend in standby mode.
Then again, the Momentus XT doesn’t look to be particularly power hungry. The drive’s power consumption is rated at just 1.6W while seeking and half that at idle. We’ll see how the drive’s real-world power draw measures up a little later in the review.
Although our focus today is on the 500GB Momentus XT, the drive is available in 320 and 250GB capacities, as well. Seagate’s official suggested retail prices for the 500, 320, and 250GB drives are $156, $132, and $113, respectively. Street prices are already lower, with Newegg listing the 500GB drive at $130 and Amazon selling the $320GB for $116 and the 250GB for $108. Even at those prices, you’re looking at a hefty mark-up over the Momentus 7200.4, whose 500GB flavor can be had for only $90.
We’ll see how the XT’s performance compares to that of the 7200.4 in a moment. First, I should point out that the XT gets two years more warranty coverage than standard Momentus drives and most other mobile hard drives on the market. Consumer-grade hard drives and the majority of SSDs are typically covered with three-year warranties. Seagate reserves five years of coverage for its enterprise-class products and flagship XT models, like this one.
Our testing methods
If you’re unfamiliar with The Twins, our new duo of storage test platforms, I recommend checking out this page from our recent VelociRaptor VR200M review. These systems pack potent hardware and have been furiously testing hard drives and SSDs for weeks now. Unfortunately, Intel still hasn’t resolved the performance scaling issue we found in its latest storage controller drivers for the P55 chipset. As a result, The Twins are still running the Microsoft AHCI driver built into Windows 7.
Before dipping into pages of benchmark graphs, let’s set the stage with a quick look at the players we’ve assembled for comparison. Seagate bills the Momentus XT as a “Premium Performance” product destined for everything from laptops to high-end small-form-factor PCs, which nicely leaves the door open for all sorts of interesting comparisons. We’ve thrown the XT into the mix against a diverse collection of alternatives, including a selection of desktop hard drives, traditional notebook drives, and a cubic assload of solid-state goodness. Below is a chart highlighting some of the key attributes of the contenders we’ve lined up.
|Flash controller||Interface speed||Spindle speed||Cache size||Platter capacity||Total capacity|
|Agility 2|| SandForce
|Caviar Black 2TB||NA||3Gbps||7,200 RPM||64MB||500GB||2TB|
|Force F100|| SandForce
|Momentus 7200.4||NA||3Gbps||7,200 RPM||16MB||250GB||500GB|
|Momentus XT||NA||3Gbps||7,200 RPM||32MB||250GB||500GB|
|PX-128M1S||Marvell Da Vinci||3Gbps||NA||128MB||NA||128GB|
|Scorpio Black||NA||3Gbps||7,200 RPM||16MB||160GB||320GB|
|Scorpio Blue||NA||3Gbps||5,400 RPM||8MB||375GB||750GB|
|SiliconEdge Blue||JMicron JMF612||3Gbps||NA||64MB||NA||256GB|
|SSDNow V+||Toshiba T6UG1XBG||3Gbps||NA||128MB||NA||128GB|
|VelociRaptor VR150M||NA||3Gbps||10,000 RPM||16MB||150GB||300GB|
|VelociRaptor VR200M||NA||6Gbps||10,000 RPM||32MB||200GB||600GB|
|Vertex 2|| SandForce
|X25-M G2||Intel PC29AS21BA0||3Gbps||NA||32MB||NA||160GB|
Given the lofty cost per gigabyte of SSDs, traditional notebook drives are the XT’s closest direct competitors. Naturally, we’ve included the old Momentus 7200.4, which preceded the XT as Seagate’s premiere mobile drive. The 7200.4 has the same 250GB platters and an identical spindle speed, but half the cache and no flash. We’ve also thrown in a couple of Western Digital’s most relevant 2.5″ mobile drives. The Scorpio Blue 750GB is the company’s highest-capacity 5,400-RPM model with a standard 9.5-mm drive height, while the Black is the company’s top-of-the-line 7,200-RPM offering. Western Digital doesn’t yet have a 7,200-RPM notebook drive with a 500GB capacity, so we’re stuck with a 320GB Black.
On the SSD front, we’ve collected all the relevant players, including drives based on Indilinx, Intel, JMicron, Marvell, SandForce, and Toshiba controllers. All the bases are covered with the exception of Crucial’s RealSSD C300, which is powered by a new Marvell controller with a 6Gbps SATA interface. Crucial just recently released a firmware update to address serious performance issues, and now that it’s out, we’ll have a full review of the RealSSD soon.
Although it might not seem like a fair fight, we’ve thrown in results for a striped RAID 0 array built using a pair of Intel’s X25-V SSDs. The X25-V only runs a little more than $100 online, and with many new gaming notebooks sporting dual hard drive bays, a mobile RAID array isn’t out of the question. Our X25-V array was configured using Intel’s P55 storage controller, the default 128KB stripe size, and the company’s latest 188.8.131.524 Rapid Storage Technology drivers.
Last, but not least, we’ve included performance data from a trio of mechanical hard drives. Western Digital’s 10k-RPM VelociRaptor VR200M is the fastest mechanical hard drive that plugs into a Serial ATA interface, and with a 2.5″ form factor (albeit one that’s too thick for standard notebooks), it’s an intriguing alternative for folks building small-form-factor rigs. Of course, plenty of SFF cases feature 3.5″ drive bays, so we’ve thrown in the fastest 7,200-RPM desktop drive on the market, Western Digital’s Caviar Black 2TB.
The block-rewrite penalty inherent to SSDs and the TRIM command designed to offset it both complicate our testing somewhat, so I should explain our SSD testing methods in greater detail. Before testing the drives, each was returned to a factory-fresh state with a secure erase, which empties all the flash pages on a drive. Next, we fired up HD Tune and ran full-disk read and write speed tests. The TRIM command requires that drives have a file system in place, but since HD Tune requires an unpartitioned drive, TRIM won’t be a factor in those tests.
After HD Tune, we partitioned the drives and kicked off our usual IOMeter scripts, which are now aligned to 4KB sectors. When running on a partitioned drive, IOMeter first fills it with a single file, firmly putting SSDs into a used state in which all of their flash pages have been occupied. We deleted that file before moving onto our file copy tests, after which we restored an image to each drive for some application testing. Incidentally, creating and deleting IOMeter’s full-disk file and the associated partition didn’t affect HD Tune transfer rates or access times.
Our methods should ensure that each SSD is tested on an even, used-state playing field. However, differences in how eagerly an SSD elects to erase trimmed flash pages could affect performance in our tests and in the real world. Testing drives in a used state may put the TRIM-less Plextor SSD at a disadvantage, but I’m not inclined to indulge the drive just because it’s using a dated controller chip.
With few exceptions, all tests were run at least three times, and we reported the median of the scores produced. We used the following system configuration for testing:
We used the following versions of our test applications:
- WorldBench 6
- Intel IOMeter 2006.07.27
- Xbit Labs File Copy Test 0.3
- HD Tune 4.01
- Visual Studio 2008 with 03-23-2010 Firefox source
- Call of Duty: Modern Warfare 2
- Crysis Warhead
The test systems’ Windows desktop was set at 1280×1024 in 32-bit color at a 75Hz screen refresh rate. Vertical refresh sync (vsync) was disabled for all tests.
Most of the tests and methods we employed are publicly available and reproducible. If you have questions about our methods, hit our forums to talk with us about them.
We’ll kick things off with HD Tune, which replaces HD Tach as our synthetic benchmark of choice. Although not necessarily representative of real-world workloads, HD Tune’s targeted tests give us a glimpse of a drive’s raw capabilities. From there, we can explore whether the drives live up to their potential.
The Momentus XT’s read speeds are disappointingly sluggish. Not only does the hybrid fail to match the performance of the much cheaper 7200.4, but it also has a slower average read speed than the 5,400-RPM Scorpio Blue. The desktop drives and SSDs are way ahead in this test.
If you look up at the line graph, you’ll see that the XT’s performance drops from nearly 100MB/s to just 75MB/s across the first 20% of the drive. The 7200.4’s read speed only falls to 94MB/s over the same interval.
HD Tune’s write speed tests take the XT’s flash memory out of the equation, which seems to suit the hybrid just fine. This time around, the new Momentus is a little bit faster than the 7200.4. The XT is even quicker than a couple of our SSD configs.
Unlike in the read speed tests, we don’t see a precipitous drop in the XT’s performance over the first portion of the drive. Writes never touch the flash, suggesting that the read-speed drop-off is tied to Adaptive Memory.
Next up: some burst-rate tests that should test the cache speed of each drive. We’ve omitted the X25-V RAID array from the following results because it uses a slice of system memory as a drive cache.
The Momentus XT appears to have a faster DRAM cache than the 7200.4. Seagate’s hybrid offers quicker burst speeds than its purely mechanical forebear, although the XT isn’t fast enough to catch the Scorpio Black.
Our HD Tune tests conclude with a look at random access times, which the app separates into 512-byte, 4KB, 64KB, and 1MB transfer sizes.
We didn’t observe the Momentus XT’s performance improve from one run to the next in HD Tune’s sustained or burst speed tests. However, Adaptive Memory quickly, er, adapted to the app’s access time tests. It only took a couple of runs for the XT’s access times to stabilize, putting the drive within striking distance of our collection of SSDs with transfer sizes up to 64KB.
The 1MB transfer size raises the access times of our SSDs by an order of magnitude, and it’s not kind to the Momentus XT, either. With this large transfer size, the XT finds itself near the back of the field with the mechanical notebook drives. How pedestrian.
With its flash memory exclusively dedicated to reads, the Momentus XT behaves like a traditional hard drive in HD Tune’s write access time tests. Unfortunately, that means much slower access times than the SSDs. The XT also falls behind the Scorpio Black at the three smallest transfer sizes, although the hybrid stays a few steps ahead of the 7200.4. I expect the XT’s larger cache deserves most of the credit on that front.
File Copy Test
Since we’ve tested theoretical transfer rates, it’s only fitting that we follow up with a look at how each drive handles a more realistic set of sequential transfers. File Copy Test is a pseudo-real-world benchmark that times how long it takes to create, read, and copy files in various test patterns. We’ve converted those completion times to MB/s to make the results easier to interpret.
Windows 7’s intelligent caching schemes make obtaining consistent and repeatable performance results rather difficult with FC-Test. To get reliable results, we had to drop back to an older 0.3 revision of the application and create our own custom test patterns. During our initial testing, we noticed that larger test patterns tended to generate more consistent file creation, read, and copy times. That makes sense, because with 4GB of system memory, our test rig has plenty of free RAM available to be filled by Windows 7’s caching and pre-fetching mojo.
For our tests, we created custom MP3, video, and program files test patterns weighing in at roughly 10GB each. The MP3 test pattern was created from a chunk of my own archive of ultra-high-quality MP3s, while the video test pattern was built from a mix of video files ranging from 360MB to 1.4GB in size. The program files test pattern was derived from, you guessed it, the contents of our test system’s Program Files directory.
Even with these changes, we noticed obviously erroneous results pop up every so often. Additional test runs were performed to replace those scores.
The Momentus XT looks very good in our file creation tests, besting the 7200.4 by huge margins across all three file sets. The XT also manages faster creation speeds than a good number of SSDs, including some very expensive SandForce-based drives that have particular problems with this test.
Impressive as it may be against the mechanical notebook drives, the Momentus XT still faces stiff competition from the desktop drives and a good number of the SSDs.
Once again, the XT struggles with sequential reads. The hybrid Momentus falls to last place with the MP3 file set, although it at least beats the two-year-old Scorpio Black with the others.
A couple of the SSDs also find the MP3 file set particularly problematic. Overall, though, the solid-state drives offer much quicker read speeds than the Momentus XT.
The XT regains its composure in the copy tests, achieving substantially better performance than our traditional notebook drives. However, the hybrid is still no match for the fastest SSDs, which complete these copy tests in less than half the time it takes the Momentus.
We’ve long used WorldBench to test performance across a broad range of common desktop applications. The problem is that few of those tests are bound by storage subsystem performancea faster hard drive isn’t going to improve your web browsing or 3ds Max rendering speeds. A few of WorldBench’s component tests have shown favor to faster hard drives in the past, though, so we’ve included them here.
Adaptive Memory does little for the hybrid’s performance in our subset of WorldBench tests. The XT didn’t get much quicker from one run to the next, and we’d even disabled the suite’s automatic defrag before each run to ensure that Adaptive Memory get hit with amnesia.
The Nero and Photoshop tests spread the field the most, and in both, the XT trails a good number of SSDs. In fact, in the Nero test, it’s even slower than the 7200.4 and Scorpio Black. The hybrid comes out ahead of both those drives in the Photoshop test, though.
Although source-code compiling isn’t a part of the WorldBench suite, we’ve often been asked to add a compile test to our storage reviews. And so we have. For this test, we built a dump of the Firefox source code from March 23, 2010 using Visual Studio 2008. This process writes over 22,000 files totaling about 840MB, so there’s plenty of disk activity. However, we had to restrict compiling to a single thread because using multiple threads in Windows 7 proved to be unstable. Mozilla recommends that Firefox be compiled with a single thread.
The storage subsystem clearly isn’t the only bottleneck in this test, but it’s interesting to note that the XT finishes the compile more than 20 seconds slower than the 7200.4. The Scorpio Black is faster, as well, but only by 16 seconds.
We’re currently looking into alternatives for this test, so if you have a suggestions for a multithreaded compiling test that will run in Windows 7, won’t be bound by our CPU, and preferably uses open-source code available to the general public, please shoot me an email.
Boot and load times
Our trusty stopwatch makes a return for some hand-timed boot and load tests. When looking at the boot time results, keep in mind that our system must initialize multiple storage controllers, each of which looks for connected devices, before Windows starts to load. You’ll want to focus on the differences between boot times rather than the absolute values.
This boot test starts the moment the power button is hit and stops when the mouse cursor turns into a pointer on the Windows 7 desktop. For what it’s worth, I experimented with some boot tests that included launching multiple applications from the startup folder, but those apps wouldn’t load reliably in the same order, making precise timing difficult. We’ll take a look at this scenario from a slightly different angle in a moment.
Boot time tests fall near the end of our test suite, so the system had already been rebooted numerous times before we conducted this testplenty of time for Adaptive Memory to optimize the flash. The Momentus XT brought our test system from POST to the Windows 7 desktop about four seconds faster than the 7200.4, making the hybrid notably quicker than our mechanical notebook drives and even a few desktop offerings. That said, one can save more seconds by moving to an SSD.
A faster hard drive is not going to improve frame rates in your favorite game (not if you’re running a reasonable amount of memory, anyway), but can it get you into the game quicker?
Adaptive memory flexes its muscles in our load time tests, although with varying degrees of success. Our first attempt at loading Modern Warfare 2‘s “O Cristo Redentor” special-ops mission took a whopping 44 seconds. The second time around, the load time dropped to 28 seconds. And the third? Just 25 seconds. With those gains, the Momentus XT easily trumps our collection of mechanical notebook drives. However, the hybrid still trails most of the SSDs by substantial margins.
In Crysis, the XT first loaded our save point from the beginning of the game in just over a minute. In the second run, the drive cut that time to 52 seconds, which is where it stayed for subsequent runs. The XT still ends up faster than our traditional notebook drives, but not by much. Once again, the SSDs are way out ahead.
TR DriveBench is a new addition to our test suite that allows us to record the individual IO requests associated with a Windows session and then play those results back on different drives. We’ve used this app to create a new set of multitasking workloads that should be representative of the sort of disk-intensive scenarios folks face on a regular basis.
Each workload is made up of two components: a disk-intensive background task and a series of foreground tasks. The background task is different for each workload, but we performed the same foreground tasks each time.
In the foreground, we started by loading up multiple pages in Firefox. Next, we opened, saved, and closed small and large documents in Word, spreadsheets in Excel, PDFs in Acrobat, and images in Photoshop. We then fired up Modern Warfare 2 and loaded two special-ops missions, playing each one for three minutes. TweetDeck, the Pidgin instant-messaging app, and AVG Anti-Virus were running throughout.
For background tasks, we used our Firefox compiling test; a file copy made up of a mix of movies, MP3s, and program files; a BitTorrent download pulling seven Linux ISOs from 800 connections at a combined 1.2MB/s; a video transcode converting a high-def 720p over-the-air recording from my home-theater PC to WMV format; and a full-disk AVG virus scan.
DriveBench produces a trace file for each workload that includes all IOs that made up the session. We can then measure performance by using DriveBench to play back each trace file. During playback, any idle time recorded in the original session is ignoredIOs are fed to the disk as fast as it can process them. This approach doesn’t give us a perfect indicator of real-world behavior, but it does illustrate how each drive might perform if it were attached to an infinitely fast system. We know the number of IOs in each workload, and armed with a completion time for each trace playback, we can score drives in IOs per second.
Below, you’ll find an overall average followed by scores for each of our individual workloads. The overall score is an average of the mean performance score in each multitasking workload.
DriveBench doesn’t produce reliable results with Microsoft’s ACHI driver, forcing us to obtain the following performance results with Intel’s 184.108.40.2064 RST drivers. We couldn’t get DriveBench to play nicely with our the X25-V RAID config, either, which is why it’s not listed in the graphs below. The app will only run on unpartitioned drives, so we tested drives after they’d completed the rest of the suite.
The Momentus XT scores a little bit higher than the 7200.4 in DriveBench, but it’s still well behind the desktop drives and even further back from the SSDs. We ran each trace five times, and performance was consistent across each test run, suggesting that Adaptive Memory isn’t of much help here.
Let’s break down the individual trace results to see if the XT pulled off any big upsets.
Nope, although the fact that the hybrid pushes more IOps than the Caviar Black with our virus scan workload is a little surprising. As you can see, the solid-state drives completely dominate our multitasking workloads.
Curious to see whether removing the multitasking element of these tests would have any bearing on the standings, I recorded a control trace without a background task.
The XT remains a little faster than the 7200.4 but much slower than the SSDs.
DriveBench lets us start recording Windows sessions from the moment the storage driver loads during the boot process. We can use this capability to take another look at boot times, again assuming our infinitely fast system. For this boot test, I configured Windows to launch TweetDeck, Pidgin, AVG, Word, Excel, Acrobat, and Photoshop on startup.
These results play out a little differently than our real-world boot tests, and they don’t favor the Momentus XT. Seagate’s hybrid scores just a little bit higher than the 7200.4, putting the XT well behind the SSD leaders.
Our IOMeter workloads are made up of randomized access patterns, presenting a good test case for both seek times and command queuing. The app’s ability to bombard drives with an escalating number of concurrent IO requests also does a nice job of simulating the sort of demanding multi-user environments that are common in enterprise applications.
See that golden yellow line right at the bottom of the graphs? Yeah, that’s the Momentus XT. Solid-state drives fare exceedingly well with the randomized access patterns that make up our IOMeter workloads, skewing the graphs considerably.
For the sake of completeness, I’ve provided some CPU utilization and efficiency graphs below. However, you’ll probably want to skip ahead to the next page for a look at IOMeter performance without the SSDs distorting the picture.
Much betterbut not for the Momentus XT. The hybrid’s transaction rates fail to scale up as the load increases from a single IO request to 32 concurrent ones. After 32 IOs, the XT’s transaction rates jump dramatically, but by that point the drive’s already way behind even its exclusively mechanical competition. Native Command Queuing has a queue depth of 32 IO requests, suggesting that the XT’s issues here are directly related to its NCQ implementation.
If SSDs do so well with the highly randomized access patterns of our four workloads, why is the Momentus XT scraping the bottom of the barrel? Probably because these workloads are running on the drive as a whole. With an SSD, the entire drive offers lightning-quick access times. The hybrid XT, however, can only provide SSD-like access times with up to 4GB of its available 500GB of storage capacity.
CPU utilization is a non-issue. All the drives are under half a percent with our Core i5-750.
The Momentus XT is no more or less efficient in terms of IOps per percent CPU utilization than the mechanical drives.
Noise levels were measured with a TES-52 Digital Sound Level meter 1″ from the side of the drives at idle and under an HD Tune seek load. Drives were run with the PCB facing up.
Our noise level and power consumption tests were conducted with the drives connected to the motherboard’s P55 storage controller.
I’ve consolidated the solid-state drives here because they’re all completely silent. The SSD noise level depicted below is a reflection of the noise generated by the rest of the test system, which has a passively-cooled graphics card, a very quiet PSU, and a nearly silent CPU cooler.
The Momentus XT is reasonably quiet when idling, but its seek noise level is even more impressive. Given the access times reported by HD Tune’s seek test, it looks like the app is pulling data from the hybrid’s flash rather than off the disk.
Mechanical hard drives usually support an Advanced Acoustic Management feature that allows users to adjust their seek ferocity. This feature isn’t supported by the Momentus XT, likely because it would interfere with Adaptive Memory’s ability to judge what to put into the flash.
For our power consumption tests, we measured the voltage drop across a 0.1-ohm resistor placed in line with the 5V and 12V lines connected to each drive. We were able to calculate the power draw from each voltage rail and add them together for the total power draw of the drive. Drives were tested while idling and under an IOMeter load consisting of 256 outstanding I/O requests using the workstation access pattern.
At idle, the Momentus XT’s power consumption is among the lowest we’ve ever measured. The hybrid draws fewer watts than a good number of SSDs and both of the other 7,200-RPM notebook drives. However, the XT’s power consumption is quite a bit higher under our IOMeter load. There, the XT pulls a little more juice than the 7200.4, probably because it has to power a 4GB flash module on top of everything else in the drive.
Capacity per dollar
After spending pages rifling through a stack of performance graphs, it might seem odd to have just a single one set aside for capacity. After all, the amount of data that can be stored on a hard drive is no less important than how fast that data can be accessed. Yet one graph is really all we need to express how these drives stack up in terms of their capacity, and more specifically, how many bytes each of your hard-earned dollars might actually buy.
We took drive prices from Newegg to establish an even playing field for all the contenders. Mail-in rebates weren’t included in our calculations. Rather than relying on manufacturer-claimed capacities, we gauged each drive’s capacity by creating an actual Windows 7 partition and recording the total number of bytes reported by the OS. Having little interest in the GB/GiB debate, I simply took that byte total, divided by a Giga (109), and then by the price. The result is capacity per dollar that, at least literally, is reflected in gigabytes.
In the preceding pages, we’ve seen the Momentus XT largely get smacked around by true solid-state drives. Now it’s payback time. Not only does the Momentus XT supply several times the storage capacity of the SSDs we’ve tested, but it does so with a cost per gigabyte that’s an order of magnitude lower. You do pay a premium for the XT versus traditional notebook drives, but the difference there is much smaller, particularly when the XT is compared to other 7,200-RPM models.
The Momentus XT wasn’t quite what I was expecting from hybrid hard drive. Sure, it combines mechanical platters with flash memory, but Adaptive Memory is a much more one-sided approach than the old ReadyDrive hybrid scheme introduced with Windows Vista. Seagate’s decision to use the flash exclusively as a read cache effectively eliminates any potential for improved write performance over mechanical hard drives. The company’s focus on improving performance also means that the flash isn’t used to conserve power. In fact, the XT is likely to sap more battery life than mechanical hard drives because it must power a flash memory chip in addition to rotating media.
Adaptive Memory is a new approach for Seagate, a technology that’s obviously been in development for some time given the vintage of the XT’s platters. The fact that the hybrid scheme is completely independent of the operating system is easily its most impressive attribute, one that should bring a smile to the face of MacBook users who have held off on SSD purchases due to OS X’s lack of TRIM support. But does it deliver on the performance front?
Sometimes, but not often. The Momentus XT showed flashes of potential in our random access time tests, offering SSD-like read access times for smaller transfer sizes. Adaptive Memory also cut Modern Warfare 2 load times nearly in half and shaved valuable seconds off Crysis: Warhead load times and the Windows 7 boot process. The XT fared extremely well in our file creation and copy tests, too, although that’s more a credit to the drive’s DRAM cache than its onboard flash.
While the Momentus XT offers good sequential write performance, sequential reads prove more problematic. The rapid drop in read speeds we observed in HD Tune isn’t encouraging, and neither is the drive’s poor read performance in FC-Test. Throw in a complete lack of performance scaling in IOMeter between 1 and 32 concurrent IO requests, and the XT looks to have just as many problems as it has potential.
Competition is the Momentus XT’s biggest obstacle, though. A good solid-state drive is going to be faster across the board, and as our results illustrate, even better at the things that Adaptive Memory does well. The XT’s relatively sluggish sequential read speeds also leave it vulnerable to mechanical notebook drives, particularly those with higher-density platters.
Given Adaptive Memory’s limitations, I think you’re better off with a two-drive hybrid that combines an SSD with secondary mechanical storage. Small-form-factor systems and larger gaming notebooks should be able to accommodate a second drive, and although such a configuration will set you back more than a single XT, I think it’s definitely worthwhile for performance-minded enthusiasts.
If you’re a notebook user with only a single hard drive bay, the Momentus XT starts to look more attractive, especially if that notebook is your primary PC. The only do-it-yourself hybrid option on that front would be an internal SSD with an external mechanical drive, which obviously has some pretty considerable drawbacks. A single Momentus XT is a more elegant and affordable solution. The Momentus XT slips into a single drive bay, and its 500GB of capacity will only set you back $130just a few dollars more than a 40GB X25-V SSD that’s barely large enough to house Windows 7 and a couple of games. Add in the quick load times that Adaptive Memory can facilitate and Seagate’s five-year warranty, and the Momentus XT becomes a viable option for notebook users looking for a bit of a performance boost without compromising storage capacity or their budgets.