"It takes a long time to bring excellence to maturity."

This maxim by improvisatore and Latin writer Publilius Syrus expresses our thoughts on the third-generation Raptor from Western Digital. This latest release in the Raptor series comes almost three years after the original Raptor WD360GD was launched.

The original Raptor was launched as the WD360GD in March of 2003 with an enterprise level 10,000 RPM spindle speed, SATA interface, 8-Megabyte buffer, 5.2 millisecond read seek time, and a single-platter design featuring 36 gigabytes of storage. While it shined in single-user performance, the drive did not fare well against its SCSI based competitors in the enterprise server market for which it was designed. It's already small size and complete lack of command queuing abilities delivered performance that was not on par with the SCSI based drives in the critical multi-user applications. However, due to the lack of entry level SCSI drives, a growing interest in SATA components, and a very avid computer enthusiast market, the drive was able to succeed until its replacement arrived.

The next version of the Raptor was launched as the WD740GD in December of 2003 and boasted several needed enhancements. These enhancements included the addition of another platter that increased the capacity to 74-gigabytes, 4.6 millisecond read seek time, a FDB based motor to address noise concerns, and ATA-4 tagged command queuing. While TCQ was a welcome addition and certainly improved the drive's I/O operations, it still was not a match for most SCSI drives in the enterprise market. This was primarily due to a lack of SATA controllers that fully supported TCQ and firmware as mature as the SCSI competition.

However, this drive was consistently and, in some cases, continues to be one of the fastest single-user performance drives available. The computer enthusiasts flocked to the drive due to its performance advantages, but the drive continued to have limited success in the enterprise server market. In a nod to the success of the drive in the computer enthusiast market and acceptance of its importance, Western Digital offered the drive in retail stores for a short time. Over the course of the last two years, the drive has been enhanced with minor revisions with the latest version, WD740GD-00FLC0, receiving tweaks that have improved its performance mainly in single-user applications. Although the drive's capacity is dated compared to the newer 300~500 gigabyte SATA drives, its price-to-performance factor is now attractive.

The latest version of the Raptor was launched as the WD1500ADFD in January of 2006 with a significant list of improvements. The drives still sports the 10,000 RPM spindle speed, 4.6 milli-second read seek time, and a two-platter design now at 150-gigabyte capacity. The platter's density has been increased to 75-gigabytes and equals those of its 10,000 RPM SCSI competitors. The buffer size has been doubled to 16 megabytes, matching the latest offerings from other drive manufacturers. The Marvell 88i8030 PATA to SATA bridge chip used on the previous two Raptors has been finally dropped in favor of a native SATA implementation. This native SATA implementation brings with it Native Command Queuing instead of the legacy firmware level Tagged Command Queuing. While the impact of NCQ can have a favorable impact in multi-user applications, it can create a performance penalty in single-user applications. Since the drive is still targeted towards the server market, it retains the first generation 150 MB/sec SATA interface instead of the 300 MB/sec SATA interface found in the newer consumer level drives. The reason for this extends to Western Digital's reluctance to use the newer electronics at this time in a server-level drive until they have been thoroughly tested and approved in an enterprise environment. However, it is almost a moot point as the current SATA drives available have not approached sustainable bandwidth operations anywhere near the 150 MB/sec interface.

Like the recently released Caviar RE2 WD4000YR drive that is based on the latest Raptor platform, this drive supports Time Limited Error Recovery and Rotary Acceleration Forward Feed operations. TLER allows the drive to signal the RAID host adapter in the event an error recovery process has exceeded time out parameters. This prevents false drive dropouts on the host adapter when utilizing RAID level 1 or above configurations. Although TLER is disabled by default, a utility might be available in the future from Western Digital to enable TLER. RAFF is an electronics feature that senses rotational vibration in a multi-drive setup and then compensates for it by controlling the drive head location. This feature is designed to keep read and write operations consistent, thus avoiding time consuming retries by the drive that could impact performance.

During the development of the WD1500 platform, Western Digital decided to acknowledge the importance of the computer enthusiast market that has made the Raptor series successful by developing the Raptor X model. The drive's model designation is WD1500AHFD and will be available shortly.

Thus, Western Digital will have two different models of basically the same drive in the market place. The WD1500ADFD will be primarily targeted as an entry-level alternative in the enterprise market with the WD1500AHFD being targeted to the computer enthusiast and gaming user.

The Raptor X model differs cosmetically from its parent by offering a transparent cover over the spindle and platter section of the drive. The cover was designed from a special grade of crystalline polycarbonate with a unique ESD-dissipative coating. The combination of this cover and the conductive frame design will protect the read/write heads from potentially damaging static discharge.

Let's see how the newest Raptor performs against other SATA based drives.



Feature Set and Test Setup: WD1500ADFD

Specification	Western Digital Raptor WD1500ADFD	Western Digital Raptor WD740GD-00FLC0
Capacity:	150GB	74GB
Interface:	SATA 150 MB/s	SATA 150 MB/s
Rotational Speed:	10,000 RPM	10,000 RPM
Buffer Size:	16 MB	8 MB
Average Latency:	2.99 ms (nominal)	2.99 ms (nominal)
Read Seek Time:	4.6 ms	4.5 ms
Write Seek Time:	5.2 ms (average)	5.9 ms (average)
Track to Track Seek Time:	0.4 ms (average)	0.6 ms (average)
Full Stroke Seek Time:	10.2 ms (average)	10.2 ms (average)
Transfer Rate - Buffer to Disk:	84 MB/s (sustained)	72 MB/s (sustained)
Number of Heads:	4	4
Number of Platters:	2	2
Command Queuing:	Native Command Queuing	Tagged Command Queuing
Acoustics- WD:	Idle - 29dBA Seek Mode 0 - 36dBA	Idle - 32dBA Seek Mode 0 - 36dBA
Other Features:	TLER - RAID Specific RAFF FlexPower	FlexPower

Both drives feature a 1.2 million MTBF hour rating and receive a 24-hour factory burn-in while offering a 5-year warranty. The 150GB version will run approximately $300 compared to $169 for the current 74GB drive. The Raptor X edition is currently listed in the $350 range.

Performance Test Configuration
Processor:	AMD Athlon 64 3500+ (Venice E4) utilized for all tests
Memory:	2 x 512MB OCZ PC4800 Elite Platinum Settings- DDR400 at (2-2-2-5, 1T)
Hard Drive(s):	1 x Maxtor MaXLine III 7L300S0 300GB 7200 RPM SATA (16MB Buffer) 1 x Samsung SpinPoint P Series SP2504C 250GB 7200 RPM SATA (8MB Buffer) 1 x Western Digital Raptor WD740GD-00FLC0 74GB 10,000 RPM SATA (8MB Buffer) 1 x Western Digital Raptor WD1500ADFD 150GB 10,000 RPM SATA (16MB Buffer)
System Platform Drivers:	NVIDIA Platform Driver - 6.70
Video Card:	1 x Gigabyte 6600GT (PCI Express) for all tests
Video Driver:	NVIDIA nForce 81.98 WHQL
Cooling:	Thermaltake Big Typhoon
Power Supply:	OCZ Power Stream 520
Operating System(s):	Windows XP Professional SP2
Motherboards:	Asus A8N-SLI Premium

Our test platform differs slightly from the previous version with the inclusion of the Asus A8N-SLI Premium board and the 81.98 video driver set. All drives are tested with NCQ or TCQ off unless otherwise noted. We have added the Samsung SP2504C and Maxtor MaXLine III 7L300S0 drives into our benchmark results for comparison to the Western Digital Raptor series.

AnandTech Storage Test Applications
iPEAK- Business Winstone:	A capture and test playback of all I/O operations within VeriTest's Business Winstone 2004 suite.
iPEAK- Content Creation Winstone:	A capture and test playback of all I/O operations within VeriTest's Multimedia Content Creation Winstone 2004 suite.
Synthetic Benchmarks:	WinBench 99 HD Tach 3.01 Everest Ultimate Edition 2.50 PCMark 2005 - HD Tests
Single Application Tests:	Timed tests of file copying, zipping, and unzipping operations.
Multi-Tasking Tests:	Timed tests of file zipping and data import operations.
Game Load Tests:	Half-Life, Doom3, and Command & Conquer: Generals

Our test suite consists of the standard benchmarks utilized at this time. We will be expanding our application and game benchmarks in the near future to further provide real world test results across a variety of programs.

The WD740GD-00FLA1 is listed on the charts as "Western Digital Raptor WD740GD-00FLA1 (74GB)" in the description field. The replacement drive for this model and the WD740GD-00FLA2 is the WD740GD-00FLC0. The WD740GD-00FLC0 is listed on the charts as "Western Digital Raptor WD740GD-00FLC0 (74GB)" in the description field.


Hard Disk Performance: iPEAK

With the variety of disk drive benchmarks available, we needed a means of comparing the true performance of the hard drives. The logical choice was Anand's storage benchmark first described in Q2 2004 Desktop Hard Drive Comparison: WD Raptor vs. the World. The iPeak test can be designed to measure "pure" hard disk performance, and in this case, we kept the host adaptor as consistent as possible while varying the hard drive models. The idea is to measure the performance of a hard drive with a consistent host adaptor. We utilized the NVIDIA nF4 SATA ports along with the NVIDIA IDE-SW driver to ensure NCQ and TCQ compatibility.

We played back Anand's raw files that recorded I/O operations when running a real world benchmark - the entire Winstone 2004 suite. Intel's iPEAK utility was then used to play back the trace file of all I/O operations that took place during a single run of Business Winstone 2004 and MCC Winstone 2004. The drive was formatted before each test run and a composite average of 3 tests on each drive was tabulated in order to ensure consistency in the benchmark.

iPeak gives a mean service time in milliseconds; in other words, the average time that each drive took to fulfill each I/O operation. In order to make the data more understandable, we report the scores as an average number of I/O operations per second so that higher scores translate into better performance.

The performance of the Raptor WD1500 is extremely impressive and represents a 13% increase over the Raptor 740GD in the Business Winstone test and an incredible 33% increase in the Content Creation Winstone test. Turning NCQ on results in a hit of 8% in the Business Winstone test and 9% in the Content Creation Winstone test. Although the drive maintains excellent performance with NCQ activated we highly recommend turning it off when used in a single-user environment. While the WD1500 random access times are slightly higher than the Raptor 740GD, the overall performance of this drive is consistently better. The Samsung SP2504C provides excellent results considering the bargain pricing on the drive and leads the Maxtor 7L300S0 in most of the test results.



Hard Disk Performance: Everest 2.50 / HD Tach

We are utilizing the Everest Ultimate Edition 2.50 HDD drive benchmarks for this article. We are utilizing their Disk Read Benchmark to track overall read performance on our drives.

The WD1500 leads the other drives in the average read access, random reads, and linear read categories. The Samsung SP2504c makes a surprise showing in the linear read beginning, random read, and buffered read categories by outscoring the WD740GD.

The HD Tach read performance results are consistent with the Everest benchmarks and show the Samsung drive having the top burst rates along with better performance than the Maxtor drive.



Hard Disk Performance: PCMark 2005

We introduce PCMark 2005 into our hard disk benchmark suite. We utilize HDD test suite within the program. The program utilizes the RankDisk application within the Intel iPEAK SPT suite of tools to record a trace of disk activity during usage of real world applications. These traces can then be replayed to generate performance measurements based upon the actual disk operations within the application. The HDD test suite contains 53% read and 47% write operations with each trace section utilizing varied amounts of read or write operations. Additional information about the test suite can be found in PDF format here.

Our test results are based upon the following trace runs:

Windows XP Startup: This test consists of 90% reading and 10% writes that tracks XP activities at start-up.

Application Loading: This test consists of 83% reading and 17% writes that tracks the opening and closing of the following programs.

• Microsoft Word
• Adobe Acrobat Reader
• Windows Media Player
• 3Dmark 2001SE
• Leadtek Winfast DVD
• Mozilla Internet Browser

General Hard Disk Drive Usage: This test consists of 60% reading and 40% writes that tracks application usage utilizing the following settings.

• Opening a Microsoft Word document, performing grammar check, saving and closing
• Compression and decompression using Winzip
• Encrypting and decrypting a file using PowerCrypt
• Scanning files for viruses using F-Secure Antivirus
• Playing an MP3 file with Winamp
• Playing a WAV file with Winamp
• Playing a DivX video using DivX codec and Windows Media Player
• Playing a WMV video file using Windows Media Player
• Viewing pictures using Windows Picture Viewer
• Browsing the Internet using Microsoft Internet Explorer
• Loading, playing and exiting a game with Tom Clancy's Ghost Recon

Virus Scanning: This test consists of 99.5% read operations as the test tracks the scanning activities of 600MB of files for viruses.

File Write: This test consists of 100% write activities by writing 680MB of files onto the hard disk.

The WD1500 leads the WD740GD by 11% in the total suite score and the Samsung SP2504C by 40%. While the Samsung drive posts very good write numbers, it is unable to match the read performance of the Raptors; but, once again, leads the Maxtor drive. The effect of enabling NCQ on the WD1500 results in a 2% drop in the overall score and 9% in the read intensive Virus Scanning benchmark.



Single Application Test Procedures:

1. File Zip Test - We take a 300MB file and measure the time that it takes for our system to compress it to ZIP format. We then start the test procedure with 300 (1MB) files to determine how the drive performs with multiple files.
2. File Unzip Test - Using the same methodology as the File Zip Test, we take a single 300MB Zip file and 300 (1MB) ZIP files and measure the time that it takes to uncompress each ZIP file successfully.
3. File Copy Test - We measure how long it takes for the system to copy a single 300MB file and 300 (1MB) files.

The WD1500 leads the other drives in this test scenario. The WD740GD-00FLC0 offers better scores than its predecessor. The Samsung SP2504C continues its lead over the Maxtor 7L300S0 in several of the benchmarks. However, the Maxtor scores very well in the Copy Folder tests.


Hard Disk Performance: Multi-Tasking & Game Tests

Multitasking Test Procedure

To provide a real world example of multitasking, we run Outlook and import 450MB of email messages into an account while running our benchmarking utility to zip a single 300MB file. To compare our results, we calculate the difference between the multitasked process and the single task file zip process.

Both Raptors provide excellent results due to their excellent reading algorithms while the Samsung drive performs closer to its price point in this benchmark.

Game Test Procedures

Our Game Level Loading Time tests include two of the most popular games: Doom 3 and Half-Life 2. Because of their high resolution textures and the large levels, the loading time for the levels of each game are long enough to show a difference between each drive.

We have also included an older strategy game, Command & Conquer: Generals, because of its longer level load times as well. Though the game is a couple of years old, it still proves to be a good measure of data loading performance.

We will be updating our benchmark suite in the near future.

The Raptors once again show excellent results with the WD1500 leading the WD740GD by a few points.



Hard Drive Performance: Thermal and Acoustics

We have revised our acoustics testing for this article. Our acoustic test utilizes our standard test bed components, but we installed our Thermaltake water-cooling system for the CPU and turned off the case fans to isolate as much case noise as possible during testing. Our OCZ power supply is virtually silent in these tests and we installed our fanless Gigabyte 6600GT video card to further reduce noise.

Our acoustic tests are designed to measure the decibel levels while the system is at idle and also under load while running the General Hard Disk Drive Usage benchmark within PCMark 2005. We found through trial and error that this particular benchmark produces controlled readings across a wide range of applications within the benchmark. This particular benchmark utilizes 60% reads and 40% writes within the trace playback file.

The measurements are taken at a distance of 5 millimeters from the rear and front of the drive being tested in order to minimize surrounding environmental noise. The reported measurements are based on an A-weighted decibel score that measures frequencies similar to the way that the human ear responds to sound. We take a total of 5 measurements for each test. We subtract the high and low scores and then average the middle three to arrive at our findings.

Our thermal tests continue to utilize sensor readings via the S.M.A.R.T. technology on the drives. We will move to an infrared thermometer testing procedure in the near future to supplement the current test procedures.

Our favorite drive for HTPC work, the Samsung SP2504C, leads the group in both thermals and acoustics. The drive was virtually silent throughout our testing and would make an excellent drive for silent systems.

The thermal characteristics of both Raptors were very close with the WD1500 running slightly hotter under the load conditions. While we did not report the infrared test results in this article, the heat sink fan area on the Raptors reached temperatures around 55c during the load testing. Of course, these temperatures would be reduced with the case fans running, but it is something to be aware of if you plan on utilizing the drive in a SFF case.

The acoustical characteristics of both Raptors were similar. You will notice a slight whine during the spin-up phase on each drive along with the same whine when the drive powers down. The reads are clearly audible, but do not have the same metallic noise that we noticed on the Maxtor drives. During normal operations, both drives could be heard, but the sounds emitting from the drive was almost dull compared to the Maxtor drives. The WD1500 was slightly louder during constant read access compared to the WD740GD. While not obtrusive in our opinion, the sound levels would be clearly evident in a silent system.



Final Words

The performance of the third-generation Raptor is truly impressive from several different angles. While, at this time, we did not test the unit against the SCSI competitors with which it is meant to compete, we did find the drive to offer the overall highest performance in the SATA market. Like its predecessor, this is the drive to have if single-user performance matters over absolute storage capability. In fact, having this drive as the primary OS and game unit with a larger capacity drive for general storage would be an ideal combination.

Western Digital has addressed the shortcomings of its previous models with the addition of a 16 MB buffer, NCQ, and native SATA implementation wrapped in two different packages depending upon your tastes and budget. With the upcoming transition from the parallel SCSI Ultra320 interface to the new Serial Attached SCSI interface, the drive stands a very good chance at competing successfully in the entry-level enterprise market. The SAS standard offers interoperability with the current SATA standards. In fact, both standards utilize the same physical connectors, while SAS host adapters can control SATA drives. This also offers the opportunity for the drive to become a standard in the workstation market where, historically, SCSI has competed well against SATA based upon performance to price ratios.

In our thermal and acoustic testing, the drive offered average scores, but considering the performance and design of the drive, we could live with these minor annoyances in our main system. However, if you plan on building a silent system or using this drive in a SFF case, please understand that the thermals and acoustics will be noticeable.

We did witness performance decreases of up to 9% in certain benchmarks with NCQ activated. We will further investigate the performance penalty of NCQ in the near future with our revised benchmark suite. At this time we recommend turning off NCQ if this drive will be utilized in a single-user environment.

With a $300 price tag, 150GB capacity, and enterprise market heritage, this drive is targeted to a different market. While Western Digital desperately wants the drive to succeed in the server arena, they are acutely aware of the fact that the computer enthusiast will likely be one of the main purchasers of this drive. The Raptor X looks like the drive to have for the serious case modder, but we feel the additional $50 in cost is better applied toward buying a large storage drive for your system.

What is our recommendation? If storage space is not of prime concern and your budget allows it, then buy this drive. It offers the best single-user performance of any drives that we have tested to date along with the safety of owning a drive designed for 24/7 operation.

It took a few years, but Western Digital brought true excellence to a now mature product.