Navigate: Advice Beginners BIOS Guide CPUs Links Mainboards Memory Network Storage Video/Sound Cards Contact Forum SiteMap Sponsors WebNews Home	.	.
Prices: Mainboards ABIT ASUS Chaintech Shuttle Soyo Tyan CPU Intel P4 2.4C-800 P4 2.6C-800 P4 2.8C-800 P4 3.0-800 P4 3.2-800 AMD AthlonXP XP 1700+ XP 2000+ XP 2400+ XP 2500+ XP 2700+ XP 3000+ XP 3200+ Athlon64 Athlon64 3200+ Athlon64 FX-51 Opteron Opteron 240 Opteron 242 Opteron 244 Opteron 246 Memory Corsair Crucial Kingston Mushkin OCZ Search Prices:

Test Setup

• ABIT KV8 MAX3
• AMD Athlon64 3200+
• 1 x 512 MB Mushkin Level II PC 3200
• ATI RADEON 9800 Pro
• 2 x Seagate Barracuda SATA-V (RAID0)
• 1 x Maxtor DiamondMax Plus9 160 GB SATA 150
• 1 x Seagate Barracuda SATA-V
• 1 x Seagate Barracuda 7200.7 SATA
• 1 x Pioneer 12x DVD

All benchmarks were carried out using the SiliconImage 3114 controller

Effective internal performance of any drive refers to the actual media performance minus the FMT derating caused by positional and ECC data wedges and is basically what can be measured by benchmark utilities like WB99 disk inspection test, ATTO, SiSoft Sandra or SCSI Toolbox to mention a few. TCDLabs HDTach also generates similar data, however, HDTach has to be taken with a grain of salt since the measurements are based not on true sequential reads.

WB99 Drive Inspection Test, SCSI Toolbox32

       

WinBench99 drive inspection test showing the sequential read performance across the different zones from OD to ID, starting at approximately 59500 MB/sec and derating to approximately 31500 MB/sec at the innermost tracks. The numbers shown here are in line with what we are getting using SCSI Toolbox32 to get an estimate of the read / write performance that is eventually limited by the HDD's media performance. As outlined here, the highest sequential write performance can exceed the read performance through better internal optimization of the data management.

ATTO Disk Benchmark; HDTach

   

ATTO basically confirms the results from SCSI Toolbox and WB99, the effective internal Read performance maxes out slightly below 60 MB/sec whereas the effective internal Writes are running up as high as 61 MB/sec. HDTach shows the effects of cache overflows along with the known issue of misreporting the write performance. The average random access time has little or no bearing on performance either and the burst read is limited primarily by the host system's chipset.

To put things a bit more in perspective, here are the sequential transfer numbers at the OD for the DiamondMax Plus9, the Seagate Barracuda 7200.7 and the Seagate SATA-V, all three of which are mainstream drives. The rationale to show the OD performance only is that of all sequential transfer data, this is the only one that matters for real world performance, whereas the "Average" sequential performance is an arbitrary definition and has no bearing on the speed of the drive.

The DiamondMax Plus9 takes a healthy performance lead over the two Seagate drives when it comes to raw linear performance at the fastest tracks. Keep in mind, however, that we are looking at "Peak" performance numbers generated by HDTach, which, in the case of the DiamondMax Plus9 are about 8MB/sec higher than the effective sequential read performance. A more realistic metric would be to average the first e.g. 20 datapoints from the log file, in that case, the performance would drop down to ~ 59MB/sec

Whatever we have looked at so far has very little to do with the real performance of the HDD, nor with MaxBoost and its performance enhancing features and we'll get to that part next.

next page:    => Basic Performance II: Maxboost =>