SSHD is the latest buzzword in the PC storage industry. Although the branding is fresh, the term refers to a class of products that’s been around for quite some time. Otherwise known as solid-state hybrid drives, SSHDs combine mechanical platters with flash-based cache memory. They first popped up in the Windows Vista era and were designed to work with that operating system’s ReadyDrive feature. Vista wasn’t exactly popular at the time, and hybrids faded into obscurity.
Seagate revisited the idea in 2011 with the Momentus XT notebook drive. This hybrid solution was independent of the OS, instead relying on drive-level intelligence to manage the flash. All things considered, it was a pretty good compromise for notebooks. A second generation Momentus XT hybrid followed the initial model, and the third iteration of the platform debuted with the SSHD moniker earlier this year.
When the latest batch of notebook-bound SSHDs was introduced, Seagate revealed that a desktop version would soon join the family. This simply named Desktop SSHD would discard the 2.5″ form factor of previous hybrids in favor of a larger 3.5″ body packed with 2TB of 7,200-RPM mechanical storage. Naturally, we were intrigued. We’ve been bugging Seagate about the drive ever since, and we finally got our hands on one to review. Let’s see what the world’s first hybrid desktop drive can do.
Porting Adaptive Memory to the desktop
Like its notebook predecessors, the Desktop SSHD adds a small amount of flash memory to a traditional hard drive. The 8GB of onboard NAND matches the flash capacity of the current generation of notebook SSHDs. 8GB doesn’t sound like a lot, but Seagate contends that typical consumer and commercial workloads deal with relatively small data sets. The firm says it experimented with different cache sizes and found 8GB to be sufficient for most workloads. Seagate concedes that its SSHDs aren’t equipped to accelerate workloads with larger data sets, such as HD video editing and CAD work. The Desktop SSHD is designed more for everyday folks than for power users.
The onboard flash can hold 8GB in an MLC configuration with two bits per cell. However, the SSHD addresses a portion of the flash as one-bit SLC NAND, which has half the effective storage density. Seagate has repeatedly dodged our questions about exactly how much of the flash is addressed as SLC NAND and whether the SLC-to-MLC ratio is static or dynamic. We do know that the MLC NAND is reserved for caching data for reads, while the SLC zones store boot data and cache some incoming writes.
There are a couple of reasons to split caching duties between SLC and MLC NAND. Due to the verification steps required when programming flash memory, one-bit cells write much faster than their two-bit counterparts. They typically have ten times the endurance, as well. Those characteristics make SLC memory ideal for write caching. Meanwhile, MLC’s higher density allows more data to be speculatively cached for read requests. That part of the flash will receive plenty of read requests, but it will only be written when Adaptive Memory copies over frequently accessed data from the platters.
Situated on the left in the picture above, the Toshiba NAND is built on a 24-nm fab process. It should be robust enough for the long haul; Seagate says the Desktop SSD has a 200% endurance “design margin” based on five years of typical use.
There’s only one flash package, so it’s important to establish performance expectations early. Modern SSDs extract much of their speed from massive parallelism. Current controllers are at their best when addressing at least 32 NAND dies across eight memory channels. Lower-capacity SSDs can’t saturate all those parallel pathways, which is why they typically carry slower performance ratings.
The Desktop SSHD’s flash component is effectively a really small SSD. Toshiba’s part number decoder indicates that the NAND package has only two channels, while Seagate’s spec sheet pegs the flash’s average data rate at 190MB/s—only a little better than one third the average throughput of today’s fastest consumer SSDs. As far as sustained speeds are concerned, the flash isn’t that much faster than the Desktop SSHD’s mechanical component, which is rated for 158MB/s. Seagate doesn’t quote the random access time of the flash, which should be orders of magnitude quicker than the 12-millisecond response time of the platters.
Deciding which data to cache in the flash falls upon Seagate’s Adaptive Memory tech. These firmware algorithms categorize data based on how frequently it’s accessed and whether doing so from the flash would improve performance. This approach draws from low-level information about how data is laid out on the drive—intelligence that software-based caching solutions lack—and it’s completely transparent to the operating system. You don’t even have to install a driver.
Seagate says the Desktop SSHD’s Adaptive Memory algorithms are exactly the same as those used in the Laptop Thin SSHD. The skinny notebook drive has a much slower 5,400-RPM mechanical component, though. Adaptive Memory doesn’t treat the two mechanical components differently. Instead, it focuses on determining whether data is sequential or random. Sequential data stays on the platters, leaving more of the flash available for random I/O.
Intercepting writes is a relatively new addition to the SSHD’s job description; until this latest generation, Seagate’s caching scheme was limited to storing data for read requests. Caching incoming data comes with some inherent risk, but Seagate says all writes are eventually passed along to the mechanical platters. If the drive loses power unexpectedly, onboard capacitors provide enough juice to flush the write cache.
|Spindle speed||7,200 RPM|
|Average solid-state data rate||190MB/s|
|Average mechanical data rate||158MB/s|
|Average read/write seek time||< 12 ms|
|Idle power||< 3.9W|
|Warranty length||Three years|
The Desktop SSHD spins dual 1TB discs at 7,200 RPM. The platters boast an areal density of 625 Gb/in², and they’re identical to the ones used in Seagate’s current lineup of traditional desktop drives. Like those models, the hybrid has 64MB of DRAM cache and a 6Gbps Serial ATA interface.
Seagate offers two Desktop SSHD models right now. The 2TB variant we’re looking at today sells for $140, while the 1TB version is priced at $100. (The 1TB version is basically the same, just with one platter instead of two.) Those prices are $30-40 higher than the going rates for equivalent 7,200-RPM hard drives in Seagate’s stable.
The price premium at least comes with a warranty bonus. Unlike Seagate’s standard desktop drives, which have two-year warranties, the SSHDs are covered for three years. That coverage matches the warranty period of most consumer-grade SSDs.
With its circuit board screwed back into place, the Desktop SSHD doesn’t look that different from standard hard drives. We know better, and now it’s time to see if the flash cache pays off…
Lining up the competition
The Desktop SSHD 2TB is a unique beast, so we’ve dug deep into our collection of storage results to find worthy competitors. 3.5″ desktop drives are obvious rivals, and we have a stack of those in the mix.
Deskstar 7K3000 3TB
|6Gbps||64MB||7,200 RPM||411 Gb/in²|
|6Gbps||64MB||7,200 RPM||625 Gb/in²|
Desktop HDD.15 4TB
|6Gbps||64MB||5,900 RPM||625 Gb/in²|
Desktop SSHD 2TB
|6Gbps||64MB||7,200 RPM||625 Gb/in²|
Caviar Black 1TB
|6Gbps||64MB||7,200 RPM||400 Gb/in²|
Caviar Black 2TB
|6Gbps||64MB||7,200 RPM||400 Gb/in²|
VelociRaptor VR200M 600GB
You’ll want to pay particular attention to how the Desktop SSHD fares against the Barracuda 3TB, which features the same mechanical technology but lacks a solid-state cache. It’s also worth keeping an eye on the Black 4TB, which is WD’s fastest 7,200-RPM desktop drive. Our Caviar Black 2TB is a bit older and has much lower-density platters, so it’s not as directly comparable to the Desktop SSHD.
Seagate’s existing hybrids aren’t directly comparable, either—they’re 2.5″ drives meant for notebooks. Still, we’ve thrown in a couple of other SSHDs to provide some context.
|Seagate Momentus XT
|Seagate Laptop Thin
The Laptop Thin SSHD is based on the same Adaptive Memory tech as the Desktop SSHD. The Momentus XT hails from the previous generation, so it can’t cache incoming writes. That drive does, however, have a faster spindle speed than the Laptop SSHD. It will be interesting to see how the Desktop SSHD stacks up against both offerings.
We’re also curious to see how the Desktop SSHD compares to modern solid-state drives, so we’ve included results for a handful of contemporary consumer models.
|Crucial M500 240GB||256MB||Marvell 88SS9187||20nm Micron sync MLC|
|Intel 335 Series 240GB||NA||SandForce SF-2281||20nm Intel sync MLC|
|OCZ Vector 256GB||512MB||Indilinx Barefoot 3||25nm Intel sync MLC|
|Samsung 840 Series 250GB||512MB||Samsung MDX||21nm Samsung Toggle TLC|
|Samsung 840 Pro 256GB||512MB||Samsung MDX||21nm Samsung Toggle MLC|
These five drives nicely cover some of the more popular controller and NAND combinations for modern SSDs. We have representatives from the high end of the spectrum, the more affordable side, and multiple points in between. All the drives are in the 240-256GB range.
If you’re a TR regular already familiar with our storage test system and methods, feel free to skip ahead to the performance results. Apart from minor tweaks to the table below, the rest of this page is copied lazily from previous reviews.
Our test methods
We used the following system configuration for testing:
Core i5-2500K 3.3GHz
Asus P8P67 Deluxe
|Platform hub||Intel P67
|Platform drivers||INF update
|Memory size||8GB (2
Corsair Vengeance DDR3 SDRAM at 1333MHz
ALC892 with 2.62 drivers
Asus EAH6670/DIS/1GD5 1GB with Catalyst 11.7 drivers
|Hard drives||Crucial M500
256GB with MU02 firmware
Intel 335 Series 240GB with 335s firmware
OCZ Vector 256GB with 10200000 firmware
Samsung 840 Series 250GB with DXT07B0Q firmware
Samsung 840 Pro Series 256GB with DXM04B0Q firmware
Hitachi Deskstar 7K3000 3TB with MKA0A580 firmware
Seagate Barracuda 3TB with CC47 firmware
Seagate Desktop HDD.15 4TB with B660 firmware
Seagate Momentus XT 750GB with SM12 firmware
Seagate Laptop Thin SSHD 500GB with SM11 firmware
Seagate Desktop SSHD 2TB with CC43 firmware
WD Caviar Black 1TB with 05.01D05 firmware
WD Caviar Black 2TB with 01.00101 firmware
WD Red 3TB with 80.00A80 firmware
WD Red 4TB with 80.00A80 firmware
WD VelociRaptor VR200M 600GB with 04.05G04 firmware
WD VelociRaptor 1TB with 04.06A00 firmware
WD Black 4TB with 01.01L01 firmware
Corsair Professional Series Gold AX650W
Windows 7 Ultimate x64
Thanks to Asus for providing the systems’ motherboards and graphics cards, Intel for the CPUs, Corsair for the memory and PSUs, Thermaltake for the CPU coolers, and Western Digital for the Caviar Black 1TB system drives.
We used the following versions of our test applications:
- Intel IOMeter 1.1.0 RC1
- HD Tune 4.61
- TR DriveBench 1.0
- TR DriveBench 2.0
- TR FileBench 0.2
- Qt SDK 2010.05
- MiniGW GCC 4.4.0
- Duke Nukem Forever
- Portal 2
Some further notes on our test methods:
To ensure consistent and repeatable results, the SSDs were secure-erased before almost every component of our test suite. Some of our tests then put the SSDs into a used state before the workload begins, which better exposes each drive’s long-term performance characteristics. In other tests, like DriveBench and FileBench, we induce a used state before testing. In all cases, the SSDs were in the same state before each test, ensuring an even playing field. The performance of mechanical hard drives is much more consistent between factory fresh and used states, so we skipped wiping the HDDs before each test—mechanical drives take forever to secure erase.
We run all our tests at least three times and report the median of the results. We’ve found IOMeter performance can fall off with SSDs after the first couple of runs, so we use five runs for solid-state drives and throw out the first two.
Steps have been taken to ensure that Sandy Bridge’s power-saving features don’t taint any of our results. All of the CPU’s low-power states have been disabled, effectively pegging the 2500K at 3.3GHz. Transitioning in and out of different power states can affect the performance of storage benchmarks, especially when dealing with short burst transfers.
The test systems’ Windows desktop was set at 1280×1024 in 32-bit color at a 75Hz screen refresh rate. 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.
The acoustic footprint of a hard drive has become one of its most important attributes—especially for PC enthusiasts who have built themselves near-silent systems. We’re a little OCD here at TR, so we’ve constructed a Box ‘o Silence to test the noise emitted by mechanical hard drives. This 18″ x 20″ anechoic chamber is lined with acoustic foam, and we suspend hard drives inside it, exactly 4″ away from the tip of our TES-52 digital sound level meter. You can read more about the setup here.
To ensure the lowest possible ambient noise levels, we swapped the test system’s graphics card for a passively cooled Gigabyte model and unplugged one of the Frio CPU cooler’s dual fans. Noise levels were measured after one minute of idling at the Windows desktop and during an HD Tune seek test.
We’ve color-coded the results by manufacturer to make the graphs easier to read. Because they have no moving parts and are essentially silent, the SSDs are missing from the noise results. When they do appear in the graphs, the corresponding bars are greyed out to set apart what is really a different class of PC storage.
The Desktop SSHD is pretty quiet for a 7,200-RPM drive. It has lower noise levels than the Barracuda 3TB at idle and under load. Seagate’s desktop hybrid is quieter than the Black 4TB, too, and by a pretty wide margin when we take seek chatter into account.
Of course, the 5,400-RPM WD Reds are quieter overall. Slower spindle speeds tend to produce less noise. So do smaller notebook drives, which is why the Momentus XT and Laptop Thin SSHD tend to have lower noise levels than the 3.5″ drives.
Power consumption was tested under load with IOMeter’s workstation access pattern chewing through 32 concurrent I/O requests. Idle power consumption was probed one minute after processing Windows 7’s idle tasks on an empty desktop.
Power consumption isn’t a huge concern for desktop systems that plug into wall sockets and have plenty of internal cooling. Differences of a few watts won’t have a big impact on your monthly utilities bill. Still, it’s nice to know that the Desktop SSHD is relatively power efficient. Despite adding cache memory, it consumes about a watt less than the Barracuda 3TB. Of course, drives with more platters tend to draw more power; the ‘cuda has an additional terabyte to spin.
Our real-world load time tests give the Desktop SSHD an opportunity to show off its caching mojo. Before timing a couple of applications, we first have to load the OS. We can measure how long that takes by checking the Windows 7 boot duration using the operating system’s performance-monitoring tools.
The Desktop SSHD reserves a portion of its cache for data associated with the Windows boot process. This boot data is stored in the NAND’s faster SLC cells, and it’s normally loaded into the flash during the installation process. Adaptive Memory is also capable of picking up on boot data with imaged installs, which is what we used with our test system. It kicked into high gear after the initial startup, shrinking the boot time from 23 seconds down to just nine for subsequent runs.
That performance puts the Desktop SSHD a little more than a second behind the SSDs. The desktop hybrid is much faster than the mechanical drives, though. The Barracuda 3TB takes nearly twice as long to load the OS. Even the latest VelociRaptor falls a couple of seconds behind.
Level load times
We observed similar behavior from the Desktop SSHD in our level load tests, though the hybrid didn’t hit its top speed until the third run in each game. It still improved substantially after the first tests, dropping from highs of 19 seconds in Portal 2 and 14 seconds in Duke Nukem Forever.
The Desktop SSHD is two seconds behind the SSDs in these tests. Its lead over the mechanical field remains, but the gaps are smaller than in our boot test. The Barracuda 3TB, for example, is only three-to-four seconds slower than Seagate’s desktop hybrid.
HD Tune — Transfer rates
HD Tune lets us look at transfer rates across the extent of the drive, and we’ve plotted the full profiles for the mechanical drives in the line graphs below. The SSDs are fast enough to throw off the scale, so we’ve left them out. You can click the buttons below each of the line graphs to see how the Desktop SSHD compares to different classes of competitors.
The Desktop SSHD matches the sequential transfer rates of its cache-less Barracuda counterpart. That’s to be expected based on the similarities between the two drives. Full-disk tests like these don’t benefit from flash-based caching.
The SSDs are substantially faster than all of the mechanical drives, regardless of whether the hard drives have onboard flash. The differences are particularly stark in the read speed tests, but they’re still prominent with writes.
This next test measures the speed of short “burst” transfers that target the DRAM caches on traditional hard drives.
Surprisingly, the Desktop SSHD scores poorly in HD Tune’s burst speed tests. The Barracuda 3TB is well over 100MB/s faster.
The Laptop Thin SSHD doesn’t fare too well, so perhaps we’re looking at a quirk of the latest Adaptive Memory implementation. Even the older Momentus XT hybrid has much higher burst rates than the Desktop SSHD.
HD Tune — Random access times
Our next set of HD Tune tests probes random access times with various transfer sizes. We’ll start with a line graph showing all the results before moving onto bar charts that cover a couple of key transfer sizes. You can click the buttons below the graph to see how the Desktop SSHD compares to different classes of rivals.
Adaptive Memory delivers snappy read access times nearly across the board. The Desktop SSHD drifts away from the solid-state drives in the 1MB test, but even then, it’s still miles ahead of the mechanical competition.
Notice that the 4KB read access times of the SSHDs are actually close to an order of magnitude slower than those of the SSDs. Those differences look tiny because the hard drives stretch the scale.
Although the Desktop SSHD’s flash cache comes through in HD Tune’s random write tests, its impact isn’t as impressive here. The Desktop SSHD still leads the mechanical drives in the 4KB test. It’s ahead of most of them in the 1MB test, too, but the Barracuda 3TB isn’t far behind. Seagate’s caching intelligence appears to be uninterested in caching large random writes, at least when HD Tune is dishing them out.
TR FileBench — Real-world copy speeds
Concocted by resident developer Bruno “morphine” Ferreira, FileBench runs through a series of file copy operations using Windows 7’s xcopy command. Using xcopy produces nearly identical copy speeds to dragging and dropping files using the Windows GUI, so our results should be representative of typical real-world performance. We tested using the following five file sets—note the differences in average file sizes and their compressibility. We evaluated the compressibility of each file set by comparing its size before and after being run through 7-Zip’s “ultra” compression scheme.
|Number of files||Average file size||Total size||Compressibility|
The names of most of the file sets are self-explanatory. The Mozilla set is made up of all the files necessary to compile the browser, while the TR set includes years worth of the images, HTML files, and spreadsheets behind my reviews. Those two sets contain much larger numbers of smaller files than the other three. They’re also the most amenable to compression.
The SSDs were tested in a simulated used state that should be representative of their long-term performance. We didn’t simulate a used state with the mechanical drives or hybrids, which tend to offer consistent performance regardless of whether we’ve run our used-state torture test.
For the most part, Adaptive Memory has no impact on FileBench performance. The Desktop SSHD was about as fast on the first run as it was on the last—and we repeated each test five times to give the caching scheme plenty of time to adapt to each file set. The only exception was the TR test, which ran at only 30MB/s the first time around. We saw a similar improvement after the first run with the Laptop SSHD, so caching appears to deserve the credit.
The Desktop SSHD has a healthy lead over the Barracuda 3TB in the TR and Mozilla tests, suggesting that caching may also play a role in the latter. Both of those tests involve large numbers of relatively small files, which would seem more ripe for caching than larger movie, RAW, and MP3 files. In those other tests, the desktop hybrid lags a little bit behind the Barracuda 3TB but manages to beat the WD Black 4TB.
Compared to the big gray elephant in the room—the SSDs, of course—the Desktop SSHD doesn’t look as competitive. The SSDs shuffle our movie, RAW, and MP3 files much quicker than the hybrids and hard drives. Their advantage isn’t as pronounced in the Mozilla and TR tests, but it’s still substantial.
TR DriveBench 1.0 — Disk-intensive multitasking
TR DriveBench allows us to record the individual IO requests associated with a Windows session and then play those results back as fast as possible on different drives. We’ve used this app to create a set of multitasking workloads that combine common desktop tasks with disk-intensive background operations like compiling code, copying files, downloading via BitTorrent, transcoding video, and scanning for viruses. The individual workloads are explained in more detail here.
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 for each multitasking workload.
On one hand, the Desktop SSHD turns in the best performance of any of the Seagate drives. It and the previous-generation Momentus XT hybrid lead Seagate’s offerings, including the Barracuda 3TB. The problem is that almost all of the other mechanical drives score higher overall. And the SSDs are several times faster than those, further muting the Desktop SSHD’s victory over its kin.
Our individual DriveBench test results are mixed. In some instances, the Desktop SSHD performs much better than the Barracuda 3TB. In others, its I/O rate is notably lower. The SSDs are in another class regardless of the workload.
TR DriveBench 2.0 — More disk-intensive multitasking
As much as we like DriveBench 1.0’s individual workloads, the traces cover only slices of disk activity. Because we fire the recorded I/Os at the disks as fast as possible, solid-state drives also have no downtime during which to engage background garbage collection or other optimization algorithms. DriveBench 2.0 addresses both of those issues with a much larger trace that spans two weeks of typical desktop activity peppered with multitasking loads similar to those in DriveBench 1.0. We’ve also adjusted our testing methods to give solid-state drives enough idle time to tidy up after themselves. More details on DriveBench 2.0 are available on this page of our last major SSD round-up.
Instead of looking at a raw IOps rate, we’re going to switch gears and explore service times—the amount of time it takes drives to complete an I/O request. We’ll start with an overall mean service time before slicing and dicing the results.
Caching clearly helps the Desktop SSHD in DriveBench 2.0. The desktop hybrid’s mean service time is almost a millisecond faster than that of its mechanical sibling. That said, the SSHD is still behind the WD Black 4TB. The old Momentus XT hybrid has a quicker mean service time, as well.
Things could be worse. The Laptop Thin SSHD falls to last place despite having the same flash cache as its desktop counterpart. The skinny notebook drive’s slow 5,400-RPM spindle speed is probably the culprit.
Let’s slice and dice the data with a few more metrics. We’ll start by splitting mean service times between read and write requests.
The differences between the read and write mean service times of the Desktop SSHD 2TB and Barracuda 3TB suggest that caching has a much bigger impact on read performance. The gap between those drives’ read service times is much larger than the one between their write service times. In both cases, the old Momentus XT notebook drive turns in a better score.
Interestingly, the SSDs aren’t as dominant here as they have been in some tests. Sure, the solid-state drives have much lower read service times. But their mean write service times aren’t otherworldly compared to the performance of the mechanical models. Some SSDs, like the Crucial M500 240GB, actually struggle a bit.
There are tens of millions of I/O requests in this trace, so we can’t easily graph all the service times to look at the variance. However, our analysis tools do report the standard deviation, which can give us a sense of how much service times vary from the mean.
There’s good news and bad news here. The good news is that the Desktop SSHD’s read service times are more consistent than those of every other mechanical drive short of the 1TB VelociRaptor. The bad news is that there’s more variance in the hybrid’s write service times than with any of the Hitachi or WD drives. At least the Desktop SSHD’s standard deviation with writes is lower than it is for the Barracuda 3TB.
We can’t easily graph all the service times recorded by DriveBench 2.0, but we can sort them. The graphs below plot the percentage of service times that fall below various thresholds. You can click the buttons below the graphs to see how the Desktop SSHD compares to different classes of mechanical and solid-state drives.
The service time distributions don’t paint the Desktop SSD in a particularly positive light. Seagate’s desktop hybrid squeezes fewer I/O requests under each service time threshold than not only its 7,200-RPM rivals, but also slower 5,400-RPM drives.
All the lines converge as we reach the right side of the graphs, indicating that the overwhelming majority of requests are serviced in 100 milliseconds or less. Service times longer than that can be sluggish enough to be perceptible, so we’ve graphed them separately. The results are expressed as a percentage of total requests. Keep in mind that our trace comprises tens of millions of requests over a nearly two-week period. Even small percentages translate to a lot of I/O.
Again, the Desktop SSHD’s flash cache seems to help more with reads than it does with writes. Although the desktop hybrid has fewer extremely long read service times than every mechanical drive except for the VelociRaptor 1TB, it stumbles with a lot more write requests than most of the other hard drives. At least the Desktop SSHD is an improvement over the Barracuda 3TB on both fronts.
Our IOMeter workloads feature a ramping number of concurrent I/O requests. Most desktop systems will only have a few requests in flight at any given time (87% of DriveBench 2.0 requests have a queue depth of four or less). We’ve extended our scaling up to 32 concurrent requests to reach the depth of the Native Command Queuing pipeline associated with the Serial ATA specification. Ramping up the number of requests also gives us a sense of how the drives might perform in more demanding enterprise environments.
There’s too much data to easily show on a single graph for each access pattern, so we’ve once again split the results by drive class. You can compare the Desktop SSHD 2TB’s performance to that of the competition by clicking the buttons below each graph. Note that the scale is different for the Raptor results.
We’ve also banished the SSDs from this set of results. Their transaction rates demand a much higher scale, making it impossible to discern what’s going on with the mechanical drives. You can see how the SSDs compare on this page of our Samsung 840 EVO review.
We normally discuss the web server results separately, since that access pattern is made up exclusively of read requests, but there’s no point here. The Desktop SSHD turns in dismal performances in all of our IOMeter tests. It pushes fewer IOps than the Barracuda 3TB at virtually every point on the scaling load in each test. The desktop hybrid does draw closer as the load increases, though.
WD’s 7,200-RPM mechanical drives have much higher I/O rates than the Desktop SSHD in each test. Even the 5,400-RPM Red 4TB has an edge most of the time, and so does the old Momentus XT notebook hybrid. I suspect the punishing nature of our scaling IOMeter workloads is simply too much for Seagate’s caching tech to handle. These tests aren’t representative of typical desktop activity.
The value perspective
Welcome to our famous value analysis, which adds capacity and pricing to the performance data we’ve explored over the preceding pages. We used Newegg to price all of the drives, and we didn’t take mail-in rebates into account when performing our calculations.
First, we’ll look at the all-important cost per gigabyte, which we’ve obtained using the amount of storage capacity accessible to users in Windows.
At $140 online, the Desktop SSHD 2TB has a higher cost per gigabyte than the Barracuda 3TB. Seven cents per gig is no worse than the going rate for WD’s 7,200-RPM offerings, though, and it’s much less than what you’ll pay for one of Seagate’s notebook hybrids.
This pricing analysis nicely puts the SSDs into perspective. As fast as they are, solid-state drives cost a lot more per gig than hybrids and hard drives.
Our remaining value calculation uses a single performance score that we’ve derived by comparing how each drive stacks up against a common baseline provided by the Momentus 5400.4, a 2.5″ notebook drive with a 5,400-RPM spindle speed. This index uses a subset of our performance data described on this page of our last SSD round-up.
Despite giving the Desktop SSHD a nice boost over the Barracuda 3TB, flash-based caching isn’t enough to elevate the hybrid above the WD Black 4TB. Based on what we’ve seen over the preceding pages, we shouldn’t be surprised. The SSHD’s Adaptive Memory tech only helped in a handful of tests, and the extra performance often wasn’t enough to bring the SSHD up to par with the fastest 7,200-RPM desktop drives.
Now for the real magic. We can plot this overall score on one axis and each drive’s cost per gigabyte on the other to create a scatter plot of performance per dollar per gigabyte. The best place on the plot is the upper-left corner, which combines high performance with a low price. We’ll focus on the hybrids and mechanical drives first.
The Desktop SSHD doesn’t look too bad in this context. It’s only a little bit slower than the Black 4TB overall, and it’s slightly cheaper per gig. Meanwhile, the Barracuda 3TB has a lower performance rating and a lower price tag to match.
Focusing on such small differences seems a little silly considering the wider picture. For a better sense of the PC storage landscape, let’s bring the SSDs and VelociRaptors onto the plot.
See that little cluster of diamonds in the lower left corner? You can make out the Desktop SSHD if you squint. Don’t bother, though. What’s important here is the vast gulf between the mechanical group and the SSDs.
The Desktop SSHD is a direct translation of Seagate’s notebook hybrid tech, and in some ways, that’s a good thing. The fact that Adaptive Memory manages the cache entirely at the drive level makes the SSHD compatible with any operating system. There’s no need to install drivers, configure software, or ensure you have a compatible hardware platform; anything with a Serial ATA port should do.
The caching scheme works, too. It cut our system’s boot time nearly in half, putting the hybrid barely behind the latest SSDs. Adaptive Memory also reduced the Desktop SSHD’s game load times, and it delivered quicker access times in our real-world DriveBench simulation.
Despite coming close to SSD performance in some tests, the Desktop SSHD fared no better than traditional mechanical hard drives in others. If the cache can’t accelerate a particular workload, you’re bound by the drive’s mechanical side. At least the 7,200-RPM platters have a little oomph—in relative terms, anyway.
Even when the cache kicked into gear, the Desktop SSHD was a step behind the SSDs. Blame the itsy bitsy flash payload, which has less storage than your smartphone and much less internal parallelism than modern SSDs. Seagate has managed to improve a few key performance metrics by marrying a tiny solid-state sidekick to its standard desktop drive, though. That’s an impressive accomplishment despite the compromise involved.
In the notebook world, where you’re often bound by a single drive bay, the compromise can makes sense. Hybrids deliver a nice speed boost while providing vast amounts of storage that would be too pricey to match with an SSD. In the desktop world, it’s rare to find systems without at least a couple of drive bays. Pairing a 3.5″ desktop drive with a 2.5″ SSD is easy, and you can decide whether to run separate system and storage drives or set up your own caching arrangement. Software-based caching schemes like Intel’s Smart Response Technology can provide similar gains with discrete SSDs.
Pairing separate solid-state and mechanical drives is more expensive than getting an SSHD, but not by as much as you might expect. Seagate’s Barracuda 2TB hard drive is priced at $100 right now, and you can add a 32GB SSD for less than $50 more. You’ll want caching software or an SRT-compatible platform to manage that relationship, but the drives ring in at only $10 more than the $140 asking price attached to the Desktop SSHD.
Given the choice, I’d rather combine separate solid-state and mechanical drives. I’d even pay a premium to get a decent-sized system SSD; the extra performance is worthwhile, and low-power secondary storage solutions are quieter than the Desktop SSHD. But my power-user preferences aside, Seagate’s desktop hybrid has some appeal for instances where a dual-drive setup isn’t possible or affordable. It’s a solid entry-level desktop offering best suited to basic systems that want to add a little pep to a lot of storage with minimal hassle.