Second-generation Serial ATA hardware has been with us since 2005. That’s a long, long time ago in PC terms. Back in 2005, motherboards supporting the then-fresh Serial ATA spec used Nvidia’s nForce4 and Intel’s 955X Express core-logic chipsets. AMD and Intel were just dipping into 64-bit waters, and then only with a single core at a time. Graphics chips had discrete pixel and vertex shaders that would be considered horribly inflexible by today’s standards. Ultraportable Windows PCs started at a couple thousand dollars, not a few hundred. And you were blissfully unaware that John and Kate Gosselin even existed.
Yeah, so-called SATA II has been around for a while. A draft specification for its third-generation replacement was presented back in 2008, with the final spec published in May of this year. So what’s new? Not a whole lot, to be honest. Native Command Queuing has gained a streaming command that can facilitate isochronous transfers. The SATA-IO governing body also promises better NCQ performance thanks to a new feature that enables “host processing and management of outstanding NCQ commands.” Oh, and the host interface speed has been doubled to 6Gbpsthat works out to about 600MB/s with protocol overhead taken into account. The higher data rate is gen-three SATA’s defining feature; the standard is referred to as SATA 6Gbps.
Getting excited about a faster hard drive interface isn’t hard. However, our first taste comes in the somewhat unlikely form of a mechanical hard drive: Seagate’s Barracuda XT.
“But the Barracuda XT is Seagate’s latest flagship desktop drive,” you say. The XT’s 500GB platters spin at 7,200 RPM, and it has a healthy 64MB of speedy DRAM cache. All of that sounds very impressive, except for a few problems.
The first issue is made quite obvious in the picture above. Yes, that’s Seagate’s top-of-the-line Serial ATA hard drive. The XT couldn’t possibly look more generic, and there’s really nothing to differentiate it visually from a bottom-rung Barracuda. Style may not count for much in a product that’ll be buried inside a case, out of view of even most side-panel windows, but it should count for something on a $300 hard drive.
|Maximum external transfer rate||600MB/s|
|Maximum sustained data rate||138MB/s|
|Average rotational latency||4.2 ms|
|Random read seek time||8.5 ms|
|Random write seek time||9.5 ms|
|Spindle speed||7,200 RPM|
|Areal density||347 Gb/in²|
|Idle acoustics||2.8 bels|
|Seek acoustics||3.2 bels|
|Warranty length||Five years|
More importantly, according to Seagate’s own data sheets, the Barracuda XT can only sustain transfer rates of up to 138MB/s. That’s on the outer edge of the platter, which is the fastest part of the disk. Even short burst transfers from this edge top out at 169MB/s, the company says. The XT may be Seagate’s fastest mechanical hard drive, but its peak media transfer rates are well within the capabilities of the second-gen, 300MB/s SATA interface.
“Surely the cache is fast enough,” thou doth protest. You’d be right to assume that the XT’s onboard DRAM memory could saturate a 600MB/s Serial ATA link, especially since Seagate itself demoed a prototype 6Gbps drive pushing nearly 590MB/s back in March. Except the XT’s isn’t nearly that fast. We asked Seagate for the XT’s maximum cache burst speed and were told that it’s only 301MB/s.
Seagate can certainly lay rightful claim to having the first SATA 6Gbps hard drive on the market. However, according to the company’s own performance numbers, the underlying drive isn’t fast enough to take proper advantage of the interface’s faster data rate.
Your motherboard probably doesn’t have any SATA 6Gbps ports, anyway, making much of this discussion academic. The ‘cuda works just fine with 3Gbps SATA controllers, and this drive is more than just a test case for the 6Gbps standard. This is the crown jewel in Seagate’s desktop hard drive lineup and the first four-disk implementation of the 500GB platters that debuted in the terabyte Barracuda 7200.12.
The Barracuda XT’s platters have an areal density of 347 Gb/in², putting the drive at a bit of a disadvantage when compared with its only four-platter, two-terabyte counterpart, Western Digital’s Caviar Black 2TB. The Caviar’s 500GB platters can accommodate 400 gigabits per square inch, which is, well, more. Higher areal densities make more data available over shorter physical distances, helping to improve performance with sequential transfers. Of course, if we’re talking sequential transfers, it makes more sense to reference linear densities, which we can derive with a little simple math. The ‘cuda has a linear density of 18.6 Gb/in, while the Caviar packs in about 8% more bits, at 20 Gb/in. We’re interested to see whether the XT can overcome that disparity in our transfer-rate tests.
Like the Caviar Black, the Barracuda XT ships with a five-year warranty. Seagate used to cover all its internal hard drive products for five years, but the longer term is now restricted to its premium and enterprise-class hard drives. The 2TB XT is currently the lone premium offering in Seagate’s desktop stable, and there are no plans to extend the XT line to lower capacity points.
So what about the controller?
So if the Barracuda XT is the first hard drive with a SATA 6Gbps interface, what about an equivalent disk controller? The only one on the market that supports the new 6Gbps standard is Marvell’s 88SE9123, a chip that got off to a bit of a rocky start earlier in the year. Problems with an initial version of the chip forced motherboard makers to drop it from their first wave of P55-based offerings. The issue was apparently traced to the ATA component of the chip rather than its SATA logic, and it’s since been resolved with a new silicon revision.
Like the Barracuda XT, the 9123 is perhaps not the most appropriate poster boy for SATA 6Gbpsnot because of past indiscretions, but due to purely technical limitations. You see, the 9123’s link to its host system is a one-lane PCI Express 2.0 interface that, at best, offers 500MB/s of bidirectional bandwidtha full 100MB/s short of 600MB/s. The Marvell chip has two SATA ports, too, so dual-drive or RAID configs will have to share the bandwidthnot that it’s going to matter much to the Barracuda XT.
The 9123 has already found its way into a batch of refreshed P55 motherboards from Asus and Gigabyte. The latter has updated its entire P55 line with new revisions that feature the Marvell chip. Asus has been somewhat more conservative, rolling out the 9123 in a couple of models, including the P7P55D Premium.
In the Premium, the 9123 rides alongside the 3Gbps SATA controller embedded in Intel’s new P55 PCH, making the board a perfect candidate for our new storage test platform. However, the P55’s half-rate PCI Express 2.0 lanes only offer 250MB/s of bandwidth, forcing Asus to employ a PLX 8613 PCIe bridge chip to keep the Marvell controller properly fed. The bridge chip sits between the Marvell controller and the P55 PCH, transforming two of the P55’s 250MB/s PCIe lanes into a single 500MB/s link. This arrangement puts the 6Gbps SATA ports one more hop away from the CPU. Data must first pass through the DMI link down to the P55, over a pair of PCIe lanes to the bridge chip, and then through another PCIe link to reach the Marvell controller.
Another curiosity associated with the Marvell controller is that its latest 126.96.36.1997 drivers section off a portion of main system memory to use as a transfer cache. That seems like an odd sort of thing to do, given how aggressively Windows preemptively populates system memory on its own, but Marvell claims Intel and JMicron use the same technique in their drivers. Seagate also recommends disabling write-cache buffer flushing with the Marvell driver to make it “behave more like” the Intel one. Write-cache buffer flushing forces applications to wait for data written to a drive to make its way through the cache and onto the platter, where it will persist in the event of power loss. The switch doesn’t appear to have much impact on the Marvell driver’s performance.
We’d already gathered a full set of performance data for the XT before Seagate dropped that particular configuration recommendation on us, along with a new firmware revision for the drive that promised better performance. We re-tested the XT using the latest firmware and the recommended driver config, and our results barely budged.
Of course, we didn’t limit ourselves to testing with the Marvell controller. The Barracuda XT also worked its way through our storage test suite connected to the 3Gbps SATA controller in Intel’s P55 chipset. We left the Intel driver in its default config, which has write-cache buffer flushing enabled, contrary to Seagate’s claims.
Since Western Digital’s Caviar Black 2TB is the XT’s most natural competitor in the desktop world, we also threw it into the mix on both the Intel and Marvell controllers. That’s as many configurations as we had time to test, but you can extrapolate how the XT might fare against a small collection of other drives by checking out the results of our Caviar Black 2TB review, which will put that drive’s performance in context.
We used the following system configuration for testing:
Intel Core i7-870 2.93GHz
|CPU/chipset link||DMI (2GB/s)|
Asus P7P55D Premium
|Chipset||Intel P55 Express|
Kingston HyperX DDR3-1600 at 1333MHz
|CAS latency (CL)||9|
|RAS to CAS delay (tRCD)||9|
|RAS precharge (tRP)||9|
|Cycle time (tRAS)||24|
Via VT2020 with
Gigabyte GeForce 8600 GT 256MB with ForceWare 190.62 drivers
Seagate Barracuda XT 2TB with CC12 firmware
Western Digital Caviar Black 2TB
Windows 7 Ultimate x64
Our test system was powered by an OCZ GameXStream power supply unit.
With the exception of our power consumption and noise levels, all tests were run at least twice, with the results averaged. We used the following versions of our test applications:
- WorldBench 6 Beta 2
- Intel IOMeter v2006.07.27
- Xbit Labs File Copy Test v0.3
- HD Tach v3.01
- Far Cry 2 v1.3
- Call of Duty 4 v1.4
The test systems’ Windows desktop was set at 1280×1024 in 32-bit color at an 85Hz screen refresh rate. Vertical refresh sync (vsync) was disabled for all tests.
All 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.
WorldBench uses scripting to step through a series of tasks in common Windows applications. It then produces an overall score. WorldBench also spits out individual results for its component application tests, allowing us to compare performance in each. We’ll look at the overall score, and then we’ll show individual application results.
WorldBench’s overall score suggests that the Barracuda XT isn’t any faster than the Caviar Black when working with common desktop applications. Let’s dig into the individual application tests for a closer look.
The ‘cuda is a little slower than the Caviar in Photoshop. Note that there isn’t much difference in drive performance between Intel’s P55 3Gbps controller and the 6Gbps Marvell chip.
The results of WorldBench’s office and multitasking tests are pretty much a wash.
In Nero, however, the Barracuda notably trails the Caviar Black when the Western Digital drive is hooked up to the Marvell controller. If you’re running on the P55, though, the ‘cuda’s a little quicker than the Caviar.
Boot and load times
To test system boot and game level load times, we busted out our trusty stopwatch.
Our test system boots just a little bit faster than with the Barracuda XT than with the Caviar Black. I suspect the Marvell controller’s slower boot times are caused not by its failure to load Windows quickly, but by the controller taking its sweet time initializing connected drives. We didn’t disable the Marvell controller when testing the drives on the P55, either.
Our drive and controller configurations run neck and neck in our level load tests. The Barracuda XT isn’t any quicker here, and neither is the Marvell controller.
File Copy Test
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.
Vista’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 or 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 Vista’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 Barracuda XT and the Marvell 88SE9123 get along quite nicely in our real-world file creation tests, sweeping all three. Their cozy relationship is most evident with the MP3 test pattern, which runs a full 25MB/s faster when the XT is connected to the Marvell chip. Interestingly, the Caviar Black runs slower on the Marvell controller with two of three test patterns. The drop in performance never amounts to more than 5MB/s, though. More worrisome for Seagate is the fact that the ‘cuda is always slower than the Caviar when both are connected to the P55 chipset.
Our file read tests turn the tablesand the standings. The Barracuda XT reigns supreme again, but it only just edges out the Caviar Black. What’s more, the Seagate drive is faster on the P55 than on the Marvell controller this time around.
The best or worst of both worlds, our copy tests combine read and write operations. That combo suits the Marvellous Barracuda duo, which is just a little bit faster than the Caviar Black on the same controller. When using the P55, the XT is quicker than the Black through two of three test patterns.
IOMeter presents a good test case for both seek times and command queuing.
The Barracuda XT gets absolutely worked in IOMeter, delivering much lower transaction rates right out of the gate. The XT’s performance scales relatively poorly as the load increases, too.
So much for the SATA 6Gbps standard’s improved command queuing, at least as it’s implemented in the Marvell controller. When connected to the Marvell, both drives hit a performance wall at 32 concurrent I/O requests, which just happens to be the depth of the Native Command, er, Queue. The P55’s performance continues to scale right up to 256 outstanding I/O requests.
At least the Marvell controller doesn’t consume more CPU cycles than the Intel one. Then again, these configs are barely registering on the CPU utilization scale with our eight-thread Core i7.
We tested HD Tach with the benchmark’s full variable zone size setting.
The Caviar Black’s areal density advantage over the Barracuda XT doesn’t amount to more than 1MB/s in HD Tach’s sustained read speed test. It’s good for up to 7MB/s in the write speed test, though. The Barracuda doesn’t read any quicker when connected to the Marvell controller, but it’s a good 2MB/s slower in the write speed test.
We’d expected HD Tach’s burst speed test to be able to tell us whether the Barracuda XT’s cache is quick enough to take advantage of a 6Gbps link. Unfortunately, the Marvell controller evidently treats system memory as the drive cache for this test, which is why we’re seeing the Marvell solution churn out more than six times the peak bandwidth available in a 600MB/s Serial ATA link. The Caviar Black achieves similarly impressive burst speeds when connected to the Marvell, though.
Perhaps more relevant are the burst speed results from our P55 configs. With that controller, the Barracuda XT pulls up 24MB/s short of the Caviar Black.
Ouch. Despite the fact that the Barracuda XT has the same 7,200-RPM spindle speed as the Caviar Black, the Seagate drive’s random access time is more than 4.5 milliseconds slower. A handful of milliseconds may not sound like much, but in a modern PC, that’s a long time. We’re also talking about a 23% difference, which is one of the biggest margins we’ve seen today.
As one might expect, there isn’t much difference in CPU utilization between the drives and controllers. Ain’t pseudo-eight-core processors grand?
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 Tach seek load. Drives were run with the PCB facing up.
Our noise level and power consumption tests were both conducted with the drives connected to the P55 rather than the Marvell controller.
Although the Barracuda XT isn’t noticeably quieter than the Caviar from a couple of feet away when the two drives are at idle, the difference is night and day when they’re seeking. The Barracuda XT’s noise levels only rise by a little more than a decibel when it starts to seek, which makes even a busy XT more than a decibel quieter than an idle Caviar. When both drives are seeking, the XT is more than seven decibels quietera huge difference compounded by the frequency with which Microsoft’s latest operating systems hit the hard drive, either to populate memory preemptively or for search-indexing purposes.
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.
The Barracuda draws slightly more power than the Caviar at idle, but nearly three watts less under our demanding IOMeter load.
Support for 6Gbps Serial ATA may be the Barracuda XT’s most intriguing attribute, but the ‘cuda has little to offer the standard. By Seagate’s own admission, the drive’s mechanical platters aren’t nearly fast enough to need even a 200MB/s link, let alone the 300MB/s one provided by the second-gen SATA spec. The 64MB cache is only a smidgen faster than 300MB/s, which means it doesn’t have a hope of exploiting a 600MB/s data pipe, either. The first 6Gbps SATA controller on the marketMarvell’s 88SE9123isn’t going to give you better than a 500MB/s transfer rate given the speed of its PCI Express link, which doesn’t help matters. The Marvell controller’s DRAM caching will no doubt be controversial, as well, and the fact that its performance trails off under heavier IOMeter loads is hardly encouraging.
This isn’t a knock on the new SATA specification. I’m actually quite eager to see what happens when next-generation solid-state drives are paired with 6Gbps controllers that can actually deliver on the peak bandwidth offered by the standard. The fact that Seagate is getting used to working with 6Gbps SATA early is probably a good thing, too. But that doesn’t make the Barracuda XT’s third-gen SATA support a compelling or even particularly relevant feature.
The obvious question, then, is what’s left of the ‘cuda when you strip away all the 6Gbps nonsense? A pretty decent hard drive, as it turns out. The Barracuda offers solid performance overall and slightly faster real-world read and copy speeds than its Caviar rival. The XT is a heck of a lot quieter, too, generating less noise while seeking than the Western Digital drive does at idle.
Of course, the Caviar Black also highlights the Barracuda’s greatest weakness: abysmal random access times that I’d wager are also responsible for the ‘cuda’s poor showing in IOMeter. Slow seeks have been a hallmark of recent Seagate drives, including the two-platter Barracuda 7200.12 on which the XT is based, so this isn’t much of a surprise.
There’s also the matter of the Barracuda XT’s availability, or lack thereof. Drives were supposed to start shipping at the end of September, but XTs aren’t listed as in stock anywhere online. Seagate does expect Newegg to have drives this week, though.
Should you buy one? Not if you’re looking for the best value. Lower capacity points offer a more attractive cost per gigabyte without sacrificing much in the way of performance, and Seagate won’t be migrating the XT down to 1.5 or 1TB anytime soon. However, if $300 is burning a hole in your pocket and you’re in the market for a reasonably quiet two-terabyte drive with speedier transfer rates than sub-7,200-RPM offerings, then yes, the Barracuda XT is a fine choice. If you want the best all-around performer, though, you’re better off with the Caviar Blackand maybe some earplugs.