Western Digital’s VelociRaptor VR200M hard drive

For quite a long time, there were three performance tiers for desktop hard drives. At the bottom, 5,400-RPM models served as the sensible baseline for those simply looking for storage. 7,200-RPM drives lived one notch higher on the ladder, offering much better performance while still being practical enough for mainstream desktops. And then there was the final tier: even faster 10,000-RPM monsters that migrated over from the world of enterprise-class workstations and servers.

Enthusiasts have always had a weakness for enterprise gear. Longer ago than I’d care to admit remembering, we couldn’t get our hands on hard drives that spun their platters faster than 7,200-RPM without dipping into expensive SCSI territory. Then along came Western Digital’s Raptor. Ostensibly designed for enterprise, this 3.5″, 10k-RPM hard drive conveniently came with a Serial ATA interface that plugged into just about any new desktop motherboard on the market. The first incarnation may only have offered 37GB of storage capacity at a time when other desktop models were pushing hundreds of gigabytes, but the Raptor line eventually spawned far more practical 74 and 150GB variants. Western Digital even gave enthusiasts a nod with a special windowed edition dubbed the Raptor X.

For its last Raptor reboot, WD followed an enterprise storage trend that’s seen 3.5″ drives give way to 2.5″ models that can be more densely packed into rack-mount servers. The resulting VelociRaptor had the same footprint as traditional mobile drives, albeit with a 15-mm drive height—too tall to slip into notebooks designed for drives only 9.5 mm thick. Despite its Yoda-like proportions, the VelociRaptor doubled its predecessor’s capacity with 300GB of storage. WD even wrapped the drive in a 3.5″ sled that conveniently slid into standard enclosure bays.

New Raptor models have come out roughly every two years since the first one was launched. Impressively, each drive has set a new performance standard upon its release. Nearly two years have passed since the VelociRaptor made its official debut, and wouldn’t you know it, Western Digital has just hatched a new one. The latest VelociRaptor VR200M retains the same 2.5″ form factor as its forebear, but it’s been working out and has bulked up to an impressive 600GB.

Before you think that we’re in for another raising of the performance bar, consider that a lot has happened in the storage world over the past two years. Just a couple of years ago, solid-state drives were rare, impractical, and stupendously expensive. Today, they’re far more common, actually quite usable, and finally starting to flirt with being justifiably affordable—depending on your budget, anyway. The question we face is obvious: can the new VelociRaptor compete in this rapidly evolving storage market, or has it become, well, a dinosaur?

The VelociRaptor evolved
The latest addition to the VelociRaptor family is an evolutionary upgrade rather than a radical departure. WD hasn’t changed the form factor at all, although it has updated the IcePack drive sled since the original VelociRaptor’s launch. The first revision’s SATA connectors weren’t in the usual spot, creating problems for hot-swap bays. An auxiliary circuit board now situates the sled’s Serial ATA ports in the correct location. (Current-generation VelociRaptors have been available with this design for a while.)

Of course, if you won’t be installing the VelociRaptor in a 3.5″ bay, you can order it without the IcePack. The sled might look like a giant heatsink, but it’s not necessary to cool the drive. Remember that the VelociRaptor was designed to live in high-density servers.

Speaking of density, the VR200M’s most impressive attribute may be its 600GB storage capacity. Western Digital pushed capacity on two fronts, upgrading the old two-platter design to three platters and increasing the capacity of each from 150 to 200GB. The result is an impressive feat of miniaturization—three 200GB platters spinning at 10,000 RPM within the confines of an enclosure that’s only a few millimeters thicker than most smartphones.

Western Digital hasn’t disclosed the actual areal densities of the platters inside either the old or the new VelociRaptors. However, a little deduction and simple math can help us estimate how many more bits per square inch the VR200M squeezes onto its platter than the old VR150M. According to official spec sheets, the VR200M’s maximum sustained data rate is 145MB/s, which is 13% faster than that of the VR150M. Since the drives share a common spindle speed and platter dimensions, we can infer that the VR200M’s linear density is likewise 13% higher. If my math is right, that translates to a 28% increase in areal density, which is slightly less than the 33% jump one might have assumed looking at platter capacities alone.

The VR200M’s higher areal density pushes tracks closer together than on the VR150M, making the former quicker when seeking from track to track. However, WD quotes the same average and full-stroke seek times for both drives. The smaller bits on higher-density platters actually make maintaining fast random access times more difficult, so there’s something to be said for the VR200M holding the line here.

Western Digital has added a 32MB cache to this new VelociRaptor, doubling the old model’s 16MB cache. The 32MB cache may indeed be adequate from a performance perspective, but it still feels a little chintzy considering the fact that WD’s top-of-the-line Caviar Black drives sport 64MB caches. Even low-power Caviar Greens are available with 64MB of cache, which makes putting 32MB on a marquee VelociRaptor seem a little short-sighted.

More puzzling than the VR200M’s cache size is the drive’s swanky new 6Gbps Serial ATA interface. There might be a small chance that this latest VelociRaptor can burst data from its cache quicker than the prior SATA standard’s 300MB/s limit, but the drive’s own spec sheet confirms that it can’t sustain even half that speed. I don’t imagine the faster interface is needed at all.

Curiously, the VR200M’s dual-core drive controller is shared with Western Digital’s RE4 2TB, which is a 7,200-RPM drive with a 3Gbps SATA interface. The RE4 is an enterprise-class model, which could explain the shared controller. After all, the VR200M does offer a number of enterprise-specific features. Support for Time-Limited Error Recovery (TLER) prevents RAID arrays from marking disks as bad prematurely, should they end up chasing down errors for too long. Also, the Rotary Acceleration Feed Forward (RAFF) capability allows the VR200M to maintain performance in tightly-packed arrays by compensating for environmental vibration.

Taking a page from its recent desktop drives, Western Digital has added NoTouch ramp loading technology to the new Raptor. This feature moves an idle drive head completely off the platter rather than letting it rest on the surface. WD claims NoTouch reduces wear on the drive. In fact, the VR200M is apparently capable of enduring 600,000 head unload and unload cycles—12 times more than the VR150M. Both drives carry mean time between failure ratings of 1.4 million hours, though. Western Digital’s five-year warranty coverage for enterprise-class hard drives persists, as well.

Meet The Twins
I’ve been working on a new storage test suite what feels like ages now, but before we get into the revamped mix of tests that will greet the VelociRaptor, I should take a moment to introduce The Twins. No, I’m not talking about Mary-Kate and Ashley. Instead, I refer to the duo of identical new test systems we’ve assembled with a smattering of cutting-edge hardware.

In the hearts of these systems beat Intel Core i5-750 CPUs. The i5-750 may not have the Hyper-Threading capacity of the i7 series, but with four Nehalem-derived cores ticking at 2.66GHz, there’s still plenty of power on tap for only $200. Thanks to Turbo Boost, the i5-750 can actually push the clock speed of a single core all the way up to 3.2GHz. However, in the interests of keeping The Twins running at the exact same clock speed, we disabled Turbo Boost and set Windows 7’s power plan to High Performance, which stops throttling from lowering the CPU clock.

Keeping the i5-750s cool are a pair of SpinQ heatsinks graciously provided by Thermaltake. These towers have a unique design that stacks radiator rings around an internal blower-style fan, and the end result is nothing short of beautiful. The SpinQ has a handy control knob that adjusts the fan speed, and I’ve had no problems running the i5-750s on an open test bench with the fan spinning at its slowest setting. You can find the SpinQ selling online for as little as $54.

4GB of memory sounded about right for these systems, and OCZ sent over a couple of Platinum DDR3-1333 kits for The Twins. These modules have been rock solid running at 1333MHz with 7-7-7-20-1T timings at 1.65V. Each 4GB kit sells for about $110 online right now.

Gigabyte kicked in matched motherboards and graphics cards for our tandem of new test systems. The P55A-UD7 is the crown jewel of Gigabyte’s P55 motherboard lineup. The UD7 has built-in 6Gbps Serial ATA support via a Marvell 9128 storage controller and SuperSpeed USB 3.0 via an NEC chip. Both of those next-gen storage chips sit behind a PLX bridge chip to ensure that the P55’s half-speed PCI Express 2.0 lanes don’t bottleneck performance.

The UD7 has all the comforts one might expect from a high-end Gigabyte board, including gobs of overclocking options and weak BIOS-level fan speed controls. Flagship products like this are rarely cheap, and the UD7 costs a pretty penny at $275 online.

Gigabyte’s passively-cooled Radeon HD 4850 1GB is quite a bit more affordable at only around $140. Although it features a last-gen graphics chip, the card is completely silent. The port cluster’s inclusion of a VGA port threw me for a bit of a loop, though. Fortunately, you still get dual digital outputs: one DVI and one HDMI.

A number of our benchmarks test drives in an unformatted state or as the secondary drive in a system. For those tests, we need a primary hard drive to host the operating system. Western Digital kindly sent over a couple of its latest terabyte Caviar Blacks to fill in when necessary. These models can currently be had for as little as $110.

Two OCZ Z-Series 550W PSUs provide power for The Twins. We didn’t need anything special on the PSU front beyond something quiet and reliable, and while I can’t speak to the latter after only a few weeks of continuous testing, these units barely make a whisper. You can pick up a 550W Z-Series PSU for $90 at Newegg.

The Twins give us a double-barreled storage test platform that should nicely represent the sort of systems enthusiasts are putting together today. We have 6Gbps and USB 3.0 covered, and thanks to Windows 7, support for TRIM with solid-state drives, as well. Thanks to Gigabyte, OCZ, Thermaltake, and Western Digital for hooking us up with the necessary hardware.

Our test methods
The Twins have been testing a collection of drives furiously for the last couple of weeks in preparation for this review. Unfortunately, I had to start all over again a week ago after discovering an issue with Intel’s latest storage controller drivers for the P55 chipset. Our test systems are now running the Microsoft AHCI driver built into Windows 7, although as you’ll see, it presents another set of challenges.

I had hoped to have a broader range of comparative results in this review, but even with dual test systems, there was only time to test a few competitors. Here’s a look at some key specifications of the rivals we’ve selected to go up against the new VelociRaptor.

As you can see, the VR200M is the only drive of the bunch with a 6Gbps SATA interface. We’ve tested the drive connected to the P55’s 3Gbps SATA controller and to the 6Gbps Marvell chip running its latest 1.0.0.1027 drivers. All the other drives were tested on the P55’s SATA controller.

Naturally, we had to pit the VR200M against its VR150M predecessor. With other hard drive makers shying away from 10k-RPM SATA drives, the old VelociRaptor is actually the new drive’s most comparable rival. We’ve also thrown Western Digital’s Caviar Black 2TB into the mix to represent the 7,200-RPM crowd. The Black offers the best overall performance of any 7,200-RPM drive, and the high areal density of its 500GB platters should give the VelociRaptor a good challenge with sequential transfers.

Since the VelociRaptor faces a storage market flooded with solid-state drives, we’ve added a couple of SSDs to the mix. Intel’s X25-M G2 160GB has been a favorite of ours since its release, and we’ve tested the drive with its latest 02HD firmware. Also included is Corsair’s Nova V128. The Nova is a new spin on Indilinx’s popular Barefoot controller, and we’ll be taking a closer look at it in an upcoming SSD round-up. As far as performance is concerned, we expect the Nova to be roughly representative of other Indilinx-based drives.

I know what you’re thinking: but what about newer SSDs like Crucial’s RealSSD C300 and all that new SandForce-based hotness that was on display at CES? The only SandForce-based model you can actually buy at the moment is OCZ’s limited-edition Vertex LE, which the company tells me is nearly sold out already. We’ll be taking a closer look at SandForce-based drives when mass-market models arrive. We’ll also be testing the RealSSD C300 but are currently waiting on Crucial to finish a firmware update that’s supposed to address some performance issues with the drive. Expect these and other SSDs, including fresh entries from Intel, Kingston, Plextor, and Western Digital, to appear in an upcoming article.

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 methods with respect to the Nova and X25-M in greater detail. Before testing the drives, each was returned to a factory-fresh state with a secure erase. 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.

According to Microsoft, the TRIM command is invoked whenever files are deleted. Flash pages housing data that have been deleted are marked as available, but it’s up to the SSD to decide when to actually erase those pages. A solid-state drive may clear pages at its leisure and is likely to do so in conjunction with subsequent writes and its own garbage-collection and wear-leveling algorithms. 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.

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.

HD Tune
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 Tach’s targeted tests give us a glimpse of each drive’s raw capabilities. From there, we can explore which drives live up to their potential.

HD Tune lets us capture transfer rates across the entire length of the disk for some additional graphing goodness. As you can see, the VR200M offers a healthy boost in read performance versus the other mechanical drives.

The SSDs are in another class entirely, however. Both offer substantially higher average read speeds, and unlike mechanical drives, they can sustain that performance across their entire capacity. At least in theory, every flash chip in an SSD is just as fast as the one sitting next to it. With mechanical drives, the outer regions of each platter offer higher sustained throughput than the inner ones.

That theory goes out the window a little with writes, which have always been a weakness for SSDs. The Nova’s write speeds oscillate in a sawtooth pattern, but it still manages to keep ahead of the VR200M, particularly as we move inward on the VelociRaptor’s platters. That said, the VelociRaptor offers much better write performance than the X25-M.

The new VelociRaptor also handily beats its predecessor and the Caviar Black here, which is to be expected. I didn’t anticipate seeing the VR200M post slower write speeds on the 6Gbps Marvell controller, though. As our nifty line graph illustrates, the Marvell controller starts out slower and never manages to catch up.

Next up: some burst-rate tests that should test the cache speed of each drive.

The 6Gbps controller redeems itself somewhat in HD Tune’s burst tests. However, the VR200M is still a long way from eclipsing the 300MB/s limit imposed by the old Serial ATA spec.

I’d elaborate on these numbers a little more, but I’m not convinced that Microsoft’s AHCI drivers are letting the drives burst as fast as they can. When initially testing with Intel’s latest AHCI drivers, HD Tune reported burst speeds 16-21MB/s higher for the VelociRaptors and Caviar Black. We’ve also seen the X25-M G2 burst at 241MB/s in HD Tach. Apparently, it’s quite difficult to make an AHCI driver that works properly with everything.

Our HD Tune tests conclude with a look at random access times, which the app separates into 512-byte, 4KB, 64KB, and 1MB tests.

With near-instantaneous access times, the SSDs easily outgun the VelociRaptor. However, the solid-state drives start to lose their luster as we access larger chunks of data. The Nova and X25-M slow by an order of magnitude when we move from 64KB to 1MB transfer sizes, although they’re still well outside the VelociRaptor’s reach.

The VR200M predictably leads the mechanical field, managing access times similar to the VR150M until we get to the 1MB transfer size. In that test, the new VelociRaptor is a good four milliseconds quicker than the old one—the same margin that the VR150M enjoys over the 7,200-RPM Caviar Black.

The results play out similarly on the write front, with the VR200M finding itself wedged between slower mechanical drives and faster SSDs. Interestingly, the mechanical drives have quicker access times with writes rather than reads, while the opposite is true for our SSDs. Indeed, the 1MB transfer size proves particularly problematic for the X25-M, which otherwise has the quickest access times.

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 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 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 VR200M holds its own against Intel’s fastest SSD in our collection of real-world write workloads, but it’s no match for the Indilinx-powered Nova. The VelociRaptor is about 60MB/s slower than the Nova with each file set, which is a huge margin considering the close transfer rates posted by the rest of the competition.

The VR200M deals with our MP3 test pattern much better than the SSDs. However, even though it leads the rest of the mechanical pack with the other file sets, the VelociRaptor loses big to the solid-state drives with the video and program files workloads. The X25-M and Nova are way out ahead in the video file set, and the Intel drive reads our collection of program files more than 100MB/s faster.

These copy tests combine read and write operations and then delete the newly created set of files. The VelociRaptor again finds itself trailing behind the SSDs and just ahead of the mechanical drives.

Application performance
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 performance—a faster hard drive isn’t going to improve your browsing or 3ds Max rendering speeds. But a few of WorldBench’s component tests have shown favor to faster hard drives in the past, so we’ve included a few of them here.

The VR200M finds itself at the head of the class in WorldBench’s Photoshop test. I’m not sure why the X25-M is so slow, but it was consistently sluggish through more than half a dozen test runs. Apart from that outlier, the field is bunched together pretty tightly.

WorldBench’s Nero test has always exploited faster hard drives, and it manages to run about 20 seconds quicker on the new VelociRaptor than on the old one. The SSDs are speedier still, and by greater margins.

This WinZip test doesn’t seem to care which drive we’re using. Only one second separates the field, which amounts to a wash.

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 thousand files totaling about 840MB, so there’s plenty of disk activity. However, we had to restrict compiling to a single core because using multiple cores in Windows 7 proved unstable.

The VelociRaptor looks to be about as fast as all the other drives when it comes to compiling Firefox. I’m curious to see whether performance in this test is affected when we delve into slower drives.

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 should be looking at differences in 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.

The SSDs boot our system a little quicker than the VelociRaptor, but you’re only looking at an advantage of a few seconds at best. The VR200M enjoys a similar edge over the other mechanical drives, and it’s even quicker when connected to the Marvell controller. However, I suspect that differences in drive initialization times are responsible for the VR200M’s apparent edge with 6Gbps SATA here.

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 one get you into that game quicker?

It certainly can in Modern Warfare 2, at least for the “O Cristo Redentor” special-ops mission. The VR200M looks pretty good here, besting its forebear by a second and the X25-M by three. Still, there’s no touching the Corsair Nova, which brings up the mission a full ten seconds faster than the VelociRaptor.

The Nova also holds a lead over the VelociRaptor in our Crysis Warhead test, which loads a save point from the beginning of the game. The difference between the two isn’t as substantial in this game, but it’s greater than the gap between the VR200M and the rest of the contenders.

Disk-intensive multitasking
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.

For each workload, we ran a disk-intensive task in the background while stepping through a succession of multiple foreground tasks. In each case, we varied the background task, while the suite of foreground tasks remained the same.

Those foreground tasks included loading up multiple pages in Firefox, opening, saving, and closing 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.

The background tasks included 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.

We can measure performance in DriveBench by playing back all of the IOs associated with each workload and timing how long that takes to complete. We know the number of IOs in each workload, and with a completion time, we can score each drive in IOs per second. During playback, any disk idle time recorded in the original session is ignored—IOs 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.

Below, you’ll find an overall score 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 or Marvell’s 6Gbps SATA driver, forcing us to obtain the following performance results with Intel’s 9.6.0.1014 RST drivers. The app will only run on unpartitioned drives, so we tested drives after they’d completed the rest of the suite.

Our overall DriveBench score underscores once again that the SSDs are in a different class than the new VelociRaptor. The VR200M is still notably quicker than its predecessor and even faster still than the Caviar Black, but those margins aren’t anywhere close to the leads enjoyed by our two SSDs.

The results of our individual tests are quite similar, with the background file copy being the one exception. With that workload, the mechanical drives are considerably more competitive, although only because the solid-state drives achieve much lower throughput than with the other workloads. Also worth noting: the VR200M has a larger lead over the mechanical pack with background file copying and virus scans than it does with the other workloads.

When creating these multitasking workloads, I decided to cut out the background tests completely for a control test. These control scores aren’t included in our overall average, but they do confirm that taking multitasking out of the equation doesn’t change the standings.

DriveBench also 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.

The VelociRaptor wins another one against mechanical competition—but loses out again to the SSDs.

IOMeter
IOMeter presents 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.

The highly randomized access patterns of our IOMeter loads give the SSDs a huge advantage. Near-instantaneous seeks are tough to beat.

Of course, the SSDs also consume more CPU cycles while generating those frightening transaction rates. The SSDs are no less efficient than the mechanical drives if you quantify the results in terms of IOs per percent CPU utilization, though.

IOMeter – No SSDs
Having SSDs in the mix really distorts our IOMeter graphs, so we’ve omitted them for another look at the results.

As you can see, the new VelociRaptor has a pronounced lead over the Caviar Black 2TB, which is very similar to Western Digital’s enterprise-class RE4. The new Raptor isn’t much faster than the old one here, but given the similar access times between the two, we didn’t expect it to be.

IOMeter registers less than 0.5% CPU utilization on our Core i5-750 with each mechanical drive.

Noise levels
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.

Solid-state drives are silent, so the noise levels you see for the X25-M and Nova represent the noise generated by the rest of the system. The VelociRaptor is pretty quiet at idle, adding less than five decibels to the system baseline. However, the VR200M is much louder than the original when seeking, and the chattering is nearly as loud as what’s generated by the notoriously noisy Caviar Black 2TB.

Most mechanical hard drives have an Automatic Acoustic Management (AAM) value that can be set between 128 and 254. Manipulating this setting tends not to affect idle noise levels, but it can dramatically impact seek noise and access times. To get an idea of the sort of performance and acoustic range available with our collection of mechanical drives, we’ve tested the seek noise level and random access time of each at the extremes of the AAM scale. By default, both of the VelociRaptors and the Caviar Black have their AAM values set to 254.

Although there’s little difference in seek times between the two VelociRaptors, the VR150M is definitely the quieter of the two. At its loudest, the original has the same seek noise levels as the new drive at its quietest. That result isn’t entirely unexpected given the fact that the VR200M is packing an extra platter. Drives with more platters tend to be louder than those with fewer.

Power consumption
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 new VelociRaptor consumes a little more power than the VR150M at idle, but the two are pretty even under load. Both are more power-efficient than the Caviar Black 2TB, but neither comes close to matching the frugal power draw of the two SSDs.

Capacity per dollar
After spending seven pages rifling through a stack of performance graphs, it might seem odd to have just a single chart 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. But 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, and since the VR200M isn’t for sale online yet, we had to use its suggested retail price of $329. 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.

We’ve spent the last few pages watching the VelociRaptor get smacked around by a couple of SSDs, but there’s more to this drive than just performance. With 600GB under the hood, the VR200M has a lot more capacity than current-generation SSDs, and it costs a heck of a lot less per gigabyte. The new VelociRaptor even offers a more attractive cost per gigabyte than the old VR150M.

Of course, the Caviar Black still has the best cost per gigabyte of the bunch. With two terabytes bursting from its 3.5″ seams and a sub-$300 street price, we wouldn’t have expected anything less.