We'll be comparing the Raptor X's performance with that of a slew of competitors, including some of the latest and greatest Serial ATA drives from Hitachi, Maxtor, Seagate, and Western Digital. We've also thrown in a Maxtor Atlas 10K V to explore how the new Raptor fares against something from the SCSI world.
The drives we'll be looking at differ when it comes to external transfer rates, spindle speeds, cache sizes, platter densities, NCQ support, 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||Capacity||Native Command Queuing?|
|Atlas 10K V||320MB/s||10,000RPM||8MB||74GB||300GB||No*|
|Barracuda 7200.7 NCQ||150MB/s||7,200RPM||8MB||80GB||160GB||Yes|
|Barracuda 7200.9 (160GB)||300MB/s||7,200RPM||8MB||160GB||160GB||Yes|
|Barracuda 7200.9 (500GB)||300MB/s||7,200RPM||16MB||125GB||500GB||Yes|
Note that the Atlas 10K V, Caviar SE16, and Raptor WD740GD lack support for Native Command Queuing. As we've mentioned, the WD740GD does support a form of command queuing known as Tagged Command Queuing (TCQ), but host controller and chipset support for TCQ is pretty thin. Our Intel 955X-based test platform doesn't support TCQ, either. Thanks to an Adaptec 29320R SCSI controller, it will support the Atlas 10K V's SCSI command queuing. The Atlas technically doesn't do NCQ, but SCSI command queuing should be just as good—if not better.
While our SCSI card supports command queuing, adding it to our test system introduces a couple of other issues. The system lacks PCI-X slots, so the card is stuck on the relatively pokey PCI bus. At the very least, this will limit the speed of burst transfers, although it could also affect performance in other tests. SCSI drives also support a WRITE_THROUGH flag that requires that data be written directly to the disk rather than to the drive cache. This feature prevents data from being lost in the event of a power failure or other interruption, but it can slow write performance. WRITE_THROUGH is an important feature for enterprise applications, so we haven't disabled it on the Atlas 10K V.
Since Seagate makes versions of the 7200.7 both with and without NCQ support, the 7200.7 in our tests appears as the "Barracuda 7200.7 NCQ" to clarify that it's the NCQ version of the drive. The Caviar RE2, Deskstar T7K250, DiamondMax 10, 7200.8, 7200.9, and Raptor X aren't explicitly labeled as NCQ drives because they're not available without NCQ support.
Our testing methods
All tests were run three times, and their results were averaged, using the following test systems.
|Processor||Pentium 4 Extreme Edition 3.4GHz|
|System bus||800MHz (200MHz quad-pumped)|
|Motherboard||Asus P5WD2 Premium|
|North bridge||Intel 955X MCH|
|South bridge||Intel ICH7R|
|Chipset drivers||Chipset 22.214.171.1243|
|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|
|Graphics||Radeon X700 Pro 256MB with CATALYST 5.7 drivers|
|SCSI card||Adaptec 20320R with 126.96.36.199 drivers|
|Hard drives|| Hitachi 7K500 500GB SATA|
Maxtor DiamondMax 10 300GB SATA
Maxtor Atlas 10K V SCSI
Seagate Barracuda 7200.7 NCQ 160GB SATA
Seagate Barracuda 7200.8 400GB SATA
Seagate Barracuda 7200.9 160GB SATA
Seagate Barracuda 7200.9 500GB SATA
Western Digital Caviar SE16 250GB SATA
Western Digital Caviar RE2 400GB SATA
Western Digital Raptor WD740GD 74GB SATA
Western Digital Raptor X 150GB SATA
|OS||Windows XP Professional|
|OS updates||Service Pack 2|
We used the following versions of our test applications:
- WorldBench 5.0
- Intel IOMeter v2004.07.30
- Xbit Labs File Copy Test v1.0 beta 13
- TCD Labs HD Tach v3.01
- Far Cry v1.3
- DOOM 3
- Intel iPEAK Storage Performance Toolkit 3.0
The test systems' Windows desktop was set at 1280x1024 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.