Western Digital’s RE3 hard drive

Manufacturer Western Digital
Model RE3 1TB
Price (MSRP) $249
Availability Now

Western Digital finally spun a terabyte up to 7,200RPM last month with the Caviar Black. With three 334GB platters backed by 32MB of cache and a five-year warranty, the Black is well equipped to compete with other terabyte offerings. This latest Caviar may not be the fastest drive on the block in a synthetic throughput drag race, but with the exception of the 10K-RPM VelociRaptor, it’s the fastest all-around drive on the market across common desktop applications, real-world file transfers, disk-intensive multitasking, and demanding multi-user loads.

The Caviar Black’s strong multitasking and multi-user performance make the drive particularly attractive for workstations and servers, but that’s enterprise territory, where single-drive implementations tend to give way to multi-drive arrays. Desktop drives can certainly be used in RAID; however, they’re not optimized to handle the error recovery quirks and additional environmental vibration associated with multi-drive arrays.

To address these enterprise challenges, Western Digital has birthed a new RE3 hard drive based on the Caviar Black. As one might expect, this third-generation RE is equipped with all sorts of RAID-specific features and firmware tweaks, but can it live up to the impressive performance standard set by its forebear? Keep reading to find out.

Caviar for enterprise

At the hardware level, the RE3 is essentially identical to the Caviar Black. Both drives use the same 334GB platters, spin them at 7,200RPM, and share a 300MB/s Serial ATA interface. However, the Caviar Black is only available in capacities between 500GB and 1TB with 32MB of cache, while the RE3 line adds a couple of lower-capacity models with 16MB caches. We’ll be looking at the terabyte RE3 today, and it’s only available with 32MB of cache.

Maximum external
transfer rate

Sustained data rate
Average read seek

Spindle speed

250, 320, 500, 750GB, 1TB

Cache size
16MB (250, 320, 500GB)
(750GB, 1TB)

Platter size

Idle power consumption
Read/write power

Idle acoustics
25-29 dBA

Seek acoustics
29-33 dBA

Warranty length
Five years

Western Digital quotes a 113MB/s sustained data rate for the RE3, which is a good 7MB/s faster than the rated sustained throughput of the Caviar Black. I wouldn’t put too much stock into that specification. WD originally claimed a 145MB/s data rate for the Caviar Black, and that’s since been revised to 106MB/s. The results of our synthetic throughput and real-world file transfer tests will give us a much better indication of the RE3’s sustained read and write rates.

Helping the RE3 keep bits moving is a dual-processor architecture inherited from the Caviar Black. The RE3 also features Western Digital’s IntelliSeek just-in-time actuator mechanism, which only moves the drive head as fast as necessary to get it into position for the next data point rather than as fast as possible. IntelliSeek can help to reduce the vibrations caused by a drive head darting back and forth at full speed—an important consideration for tightly-packed RAID arrays.

Too much environmental vibration can knock the drive head off course, slowing performance in the process. Since IntelliSeek can only reduce vibrations rather than eliminate them completely, the RE3 employs Read Ahead Fast Forward (RAFF) tech that monitors both linear and rotational vibration and adjusts the drive head’s position accordingly. Western Digital claims that this latest RAFF implementation boosts the RE3’s performance by 60% over the previous generation RE2 in “high vibration environments.”

To further protect the RE3 from environmental hazards, WD has equipped the drive with a multi-axis shock sensor similar to what you’d find in a mobile hard drive. This sensor is designed to compensate for physical shock rather than vibration, with an eye toward protecting data from a catastrophic head crash rather than maintaining consistent performance through physical trauma. WD’s NoTouch technology also makes an appearance in the RE3, moving the drive head completely off the platter when it’s not in use.

When deployed independently, hard drives are left to perform error recovery on their own while the rest of the system waits. RAID controllers aren’t keen to idle, though, and if a drive takes too long to recover from an error, it’s often marked as bad and dropped from the array. To prevent premature ejection from an array, the RE3 features Time-Limited Error Recovery (TLER) that, er, limits the amount of time the drive will be unresponsive while attempting error recovery on its own. If the drive can’t recover the error within this limited time span, it simply gives up, deferring error recovery to the RAID controller

Like most enterprise-class Serial ATA hard drives, the RE3 is rated for a Mean Time Between Failures of 1.2 million hours. The drive also receives additional validation testing that isn’t performed on the Caviar Black, although both drives are covered by the same five-year warranty.

Test notes
We’ll be comparing the performance of the RE3 with that of a slew of competitors, including some of the latest and greatest Serial ATA drives from Hitachi, Samsung, Seagate, and Western Digital. These drives differ when it comes to external transfer rates, spindle speeds, cache sizes, platter densities, and capacity, all of which can have an impact on performance. Keep in mind the following differences as we move through our benchmarks:

Max external
transfer rate

Spindle speed

Cache size

Platter size


Barracuda 7200.11
300MB/s 7,200-RPM 32MB 250GB 1TB

Barracuda ES.2
300MB/s 7,200-RPM 32MB 250GB 1TB

Caviar Black
300MB/s 7,200-RPM 32MB 334GB 1TB

Caviar GP
300MB/s 5,400-7,200-RPM 16MB 250GB 1TB

Caviar SE16 (640GB)
300MB/s 7,200-RPM 16MB 320GB 640GB

Deskstar 7K1000
300MB/s 7,200-RPM 32MB 200GB 1TB

Raptor WD1500ADFD
150MB/s 10,000-RPM 16MB 75GB 150GB

300MB/s 5,400-7,200-RPM 16MB 250GB 1TB

300MB/s 7,200-RPM 32MB 334GB 1TB

SpinPoint F1
300MB/s 7,200-RPM 32MB 334GB 1TB

VelociRaptor VR150
300MB/s 10,000-RPM 16MB 150GB 300GB

Performance data from such a daunting collection of drives can make our bar graphs a little hard to read, so we’ve colored the bars by manufacturer, with the RE3 appearing in brighter blue than the rest of Western Digital’s drives. Pay particular attention to how the RE3 stacks up against its desktop twin, the Caviar Black.

Our testing methods
All tests were run three times, and their results were averaged, using the following test system.

Processor Pentium 4 Extreme Edition 3.4GHz
System bus 800MHz (200MHz quad-pumped)
Motherboard Asus P5WD2 Premium
Bios revision 0422
North bridge Intel 955X MCH
South bridge Intel ICH7R
Chipset drivers Chipset
Memory size 1GB (2 DIMMs)
Memory type Micron DDR2 SDRAM at 533MHz
CAS latency (CL) 3
RAS to CAS delay (tRCD) 3
RAS precharge (tRP) 3
Cycle time (tRAS) 8
Audio codec ALC882D
Graphics Radeon X700 Pro 256MB with CATALYST 5.7 drivers
Hard drives Seagate Barracuda 7200.11 1TB
Seagate Barracuda ES.2 1TB
Samsung SpinPoint F1 1TB
Hitachi Deskstar 7K1000 1TB

Western Digital RE2- GP 1TB

Western Digital Caviar GP 1TB

Western Digital VelociRaptor 300GB

Western Digital Raptor WD1500ADFD 150GB

Western Digital Caviar Black 1TB

Western Digital Caviar SE16 640GB
Western Digital RE3 1TB
OS Windows XP Professional
OS updates Service Pack 2

Thanks to NCIX for getting us the Deskstar 7K1000 and Spinpoint F1.

Our test system was powered by OCZ PowerStream power supply units.

We used the following versions of our test applications:

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.

Unsurprisingly, the RE3 ties the Caviar Black’s overall WorldBench score.

MusicMatch Jukebox

Windows Media Encoder

Adobe Premiere

VideoWave Movie Creator

Adobe Photoshop

ACDSee PowerPack

Microsoft Office


Mozilla and Windows Media Encoder



As you can see, we’re skimming through the benchmarks today. Why? Because as expected, the RE3’s performance tracks very closely with that of the Caviar Black.

Boot and load times
To test system boot and game level load times, we busted out our trusty stopwatch.

The RE3’s level load times are a hair quicker than those of the Caviar Black, but well within the margin of error we’d associate with manual stopwatch timing. More interesting are the boot time results, which have the RE3 a full 10 seconds slower than its desktop counterpart. These results were consistent across multiple test runs, although boot times are of lesser importance in enterprise environments where reboots are generally avoided.

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. File copying is tested twice: once with the source and target on the same partition, and once with the target on a separate partition. Scores are presented in MB/s.

To make things easier to read, we’ve separated our FC-Test results into individual graphs for each test pattern. We’ll tackle file creation performance first.

The RE3 doesn’t betray its Caviar Black underpinnings with FC-Test’s file creation workloads.

The drives are evenly matched when we shift our attention to read performance, as well.

As they are when we combine reads and writes with a suite of copy tests.

Switching gears to a file copy from one partition to another doesn’t change the picture, either. If you’re shuffling files around, the RE3 is every bit as quick as the Caviar Black, which is arguably the fastest 7,200-RPM drive in these tests.

iPEAK multitasking
We’ve developed a series of disk-intensive multitasking tests to highlight the impact of seek times and command queuing on hard drive performance. You can get the low-down on these iPEAK-based tests here. The mean service time of each drive is reported in milliseconds, with lower values representing better performance.

Our iPEAK multitasking workloads show that, again, the RE3 offers equivalent performance to the Caviar Black. That’s a good thing.

IOMeter presents a good test case for both seek times and command queuing.

The Caviar Black showed its enterprise potential in IOMeter, and the RE3 doesn’t disappoint here. It doesn’t improve on its predecessor, either, but the fact that the RE3 matches the peak transaction rate of the 10K-RPM Raptor WD1500ADFD with three test patterns is nothing short of remarkable.

Turning our attention to IOMeter response times, the RE3’s performance continues to track with that of the Caviar Black.

Yeah. Let’s move along.

HD Tach
We tested HD Tach with the benchmark’s full variable zone size setting.

HD Tach’s suite of synthetic throughput and access time tests shows the RE3 neck-and-neck with the Caviar Black. CPU utilization results are within HD Tach’s +/- 2% margin of error for that test.

Noise levels
Noise levels were measured with an Extech 407727 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.

Low noise levels may not be vital for server applications, but they matter for workstations, particularly if you’re running multiple drives. The RE3 proves to be a little louder than the Caviar Black both at idle and under load, and that makes it one of the loudest drives of the bunch.

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. Through the magic of Ohm’s Law, we were able to calculate the power draw from each voltage rail and add them together for the total power draw of the drive.

Like the Caviar Black that preceded it, the RE3 is quite power-efficient for a terabyte hard drive. Western Digital’s GreenPower Caviar GP (now known as the Caviar Green) and RE2-GP sip fewer watts, of course, but their slower spindle speeds don’t deliver nearly the same performance punch.


The Caviar Black’s most impressive performances came with disk-intensive multitasking and demanding multi-user loads—staples of enterprise applications. In fact, when I reviewed the Black just one month ago, I said that it was “better suited to workstation and server environments that will capitalize on the drive’s quick access times and strong performance with workloads that include more randomized I/O request patterns.”

As we’ve seen, the RE3 has lost none of the magic that made the Black so impressive. Western Digital has essentially taken a desktop drive whose performance suited workstation and server workloads and wrapped it up in an enterprise-friendly package littered with RAID-specific optimizations. Read Ahead Fast Forward and Time-Limited Error Recovery may not be necessary in single-drive desktops, but they’re potentially important features for multi-drive applications that have to deal with different error recovery rules and additional environmental vibration.

One might expect Western Digital to charge a hefty premium for the RE3’s RAID-specific goodness, yet the drive actually carries the same $250 suggested retail price as the Caviar Black 1TB. Already, the Black’s street price has plunged to just $170 online, and although the RE3 isn’t yet listed in our price search engine, I’d expect it to ring in at under $250 when it becomes widely available.

If you’re looking to build a high-performance, multi-terabyte RAID array for a server or workstation, the RE3 is currently the best candidate on the market. Enthusiasts looking for a desktop drive should stick with the Black, though. Now that Western Digital has extended its five-year warranty to high-end Black series desktop drives, there’s no reason to dip into the RE line for additional coverage.

Comments closed
    • DMF
    • 12 years ago

    Of two (now three) RE3s, I had one go bad within a couple days. But I don’t mind that much so long as the MTBF of the survivors stays high.

    You’re measuring dbA at 1″ (inch)? Is that open air or hard mounted? Same mounting for all testees?

    Anecdotal, sure, but in my experience the RE3 is one quietest drives I’ve ever deployed. At least as quiet as some older Hitachi “low noise” drives I’ve used. I can’t hear them above <20 dB case fans.

    • DMF
    • 12 years ago

    Of two (now three) RE3s, I had one go bad within a couple days. But I don’t mind that much so long as the MTBF of the survivors stays high.

    You’re measuring dB at 1″ (inch)? Is that open air or hard mounted? Same mounting for all testees?

    Anecdotal, sure, but in my experience the RE3 is one quietest drives I’ve ever deployed. At least as quiet as some older Hitachi “low noise” drives I’ve used. I can’t hear them above <20 dB case fans.

    • xtalentx
    • 13 years ago

    Hopefully these hold up better than the RE2. The RE2 has a horrendous DOA rate. I am a WD fan but the RE2 made me question the direction WD is going in right now.

      • l33t-g4m3r
      • 13 years ago

      absolutely. I just had a RE2 die on me.
      you can tell they are trying to fix the problem with new protection features.

    • xastware
    • 13 years ago

    It’s nice to see you concern a big deal to confuse people with so called Caviar SE16 while on the official site this drive is specified into Caviar Blue series

    Nice review but it be better to show some other benchies than win & doom3 load, it not unimportant but these few seconds on the action we do once a day or less, are really irrelevant. Or some weird test from ipeak that are questionable by itself cause they’re synthetic and therefore easy to be manipulated, ntm that some actions combined in those tests really some of never/rarely used in real life on workstatons. and IOmeter is more reliable representation of server workload than ipeak (if that was the intention)

    • Dr_b_
    • 13 years ago

    Got 4 of these drives in 2 comps in a RAID 1 array on both systems. Very fast performers, and really I don’t notice any noise from them. they are cool and quiet. RAID 1 Intel ICH9R, and MD array in a linux box.

    Replaced 2 seagate ES2’s, and the performance gain was phenominal.

    I am really impressed with these drives, even though the cover isn’t flush on the top, I thought it looked really cheesey, but the proof is in the performance and durability and they have it (so far).

    • continuum
    • 13 years ago

    WD’s 1TB GP and RE2-GP are fixed 5400rpm. Please fix those spec charts!

      • Dissonance
      • 13 years ago

      WD has declined to reveal the exact spindle speed of its GreenPower drives, saying only that it falls between 5,400 and 7,200RPM. It’s surely closer to the former than the latter, though.

        • mattthemuppet
        • 13 years ago

        numerous sites (SPCR included) have spectrum analyses of the sound the GP drives make and they all have a frequency peak corresponding to 5400rpm. NONE of them have seen anything else, so I think it’s pretty safe to say, irrespective of WD’s fudging, they are 5400rpm drives.

        Can this finally be laid to rest?

    • stevew49
    • 13 years ago

    Are you able to measure peak power consumption during disk spin-up?

    Also, when do you start timing the boot process? It would be interesting if the RE3 is spinning up slower than the Black in order to reduce peak power consumption (it becomes big deal when you have eight of these things hanging off a single power supply, and they all decide to draw 35W at the same time).

      • Dissonance
      • 13 years ago

      Peak power draw during spin-up is just under 25W. When measuring boot times, we start the clock as soon as the power button is pressed.

        • stevew49
        • 13 years ago

        thanks for that info. Still need a 300W power supply, then.

        In the RE3 series, it appears WD has started setting the jumpers to enable “power up in standby” (PM2), which is useful if you have a RAID controller that can stagger the spin up of the attached drives, which reduces the peak current draw overall. My old WD4000YRs couldn’t do that, but as they get RMA’d with WD5002ABYS’s, suddenly I’ve noticed the difference.

        If the Black has PM2 disabled by default, and the RE3 has PM2 enabled by default, then that may explain the difference you measured in boot time. Would you be interested in switching the jumper and measuring it again?

        I’ve gone through the WD advance-RMA process *many* times over the past year, as all of the WD4000YRs in my array have failed on or about their fourth birthday. I’ve never had a problem getting replacement drives under warranty through their website, but it takes two weeks after I file a claim for them to realize that they don’t have 400GB RAID Edition drives in stock anymore, then a couple of emails back and forth for me to convince them that anything 400GB or greater will work on a Highpoint 2320, and then finally they ship the drive 2nd-day air.

    • Spotpuff
    • 13 years ago

    Weird, a review of an RE drive in non-raid…

    • MadManOriginal
    • 13 years ago

    Scott or Geoff, when are you going to have a review of the Seagate 1.5TB? I can only find one ‘pro’ review but it was really poor with very few tests and no power draw or noise measurements.

      • Dissonance
      • 13 years ago

      As soon as we can get our hands on one. We should have something in the next few weeks.

        • MadManOriginal
        • 13 years ago

        I’m mainly interested in noise levels, a little less in power draw. I need to choose between those and WD GP drives but will take performance and maximized space as long as they aren’t really noisy on seeks like a Seagate 7200.10.

        Get to it! 😉

      • l33t-g4m3r
      • 13 years ago

      I just got one.
      It’s pretty fast and not too noisy either.

    • uksnapper
    • 13 years ago

    An observation on WD 5 year warranty .
    WD would not honour the warranty on a drive that failed when the vendor I purchased from went out of business so make sure you purchase from an established company that looks as if it will still be around in 5 years

      • indeego
      • 13 years ago

      Did you talk to a CS manager when you queried? That sounds absurdg{<.<}g

        • accord1999
        • 13 years ago

        Yeah, that seems strange. In Canada, WD doesn’t even require a receipt for a warranty replacement.

    • Sumache
    • 13 years ago

    I’m curious as to how well these drives perform in a RAID array compared to the desktop counterparts.

      • not@home
      • 13 years ago

      and i would like to see a review of some hardware raid controllers. possibly put the two reviews together. i love hardware raid controllers, but mine is getting old (2001) and its PATA only. time for an upgrade.

        • Spotpuff
        • 13 years ago

        Wonder if hardware RAID 5 will ever make it to the mainstream with low cost hardware controllers.

    • shank15217
    • 13 years ago

    Enterprise storage gets most of its performance from having lots of drives and distributing the storage loads across those drives. I am not sure if disk level multi-io tasking optimization makes a difference. Maybe its better to have fast drives with larges caches and good sequential performance and let the file system and the raid controller sort out how to queue requests to the drives. It would be great if tech-report could test several drives in a raid setup. Obviously benchmarking would be difficult, probably why there are no good ones on the web. I just bought 6 x 1TB caviar black drives with an areca ARC 1260 for work. If anybody can suggest some good benchmarks let me know. The most powerful one I could get my hands on was Flexible I/O

    §[<http://freshmeat.net/projects/fio/<]§ Using this I can create most types of io patterns.

      • eitje
      • 13 years ago

      there’s no dash, dag-nabbit!

        • shank15217
        • 13 years ago

        lol, i’ll get on it 😀

    • Usacomp2k3
    • 13 years ago

    It seems that previous RE’s had a bit more of a lead compared to the consumer varieties than this one does. I guess the consumer one’s are just catching up.

Pin It on Pinterest

Share This