Presentation of competitors and testing system



These are the 5 brand new, still wrapped in their sealed covers hdd's, kindly supplied by Loveno. We'll be putting them on the testbench for this roundup:

>

Brand & type	Maxtor Diamondmax	Maxtor Diamondmax Plus9	WD Caviar 800BB	WD Caviar 800JB	Seagate Barracuda IV
Capacity	80 gb	80 gb	80 gb	80 gb	80 gb
formatted capacity	74,5 gb	76,33 gb	74,43 gb	74,53 gb	74,53 gb
Intergrated controller	ata5/udma 133	ata5/udma 133	ata5/udma 100	ata5/udma100	ata5/udma 100
Buffer size	2 mb	8 mb	2 mb	8 mb	2 mb
Average seek time	< 8 ms	9,4 ms	8,9 ms	8,3 ms	9,5 ms
price	NA	€130	€122	€133	€133,99

The system I used to run HD Tach, Sisoft Sandra and file moving tests on, is a PIV 2.26 with 512 mb ram and a well filled Seagate cuda IV, operated by Windows XP professional. I didn't bother doing a clean install on my primairy disk to come as close to real life writing/reading circumstances as possible. This implements that some testresults may be not as good as the manufacturer claims they should be ;)

The first test on the "to do list" was HD tach, this test is to be performed on a non formatted drive meaning the raw performance of the drive comes to the surface.
Next test: Sisoft Sandra 2002 , the results here are to be interpreted with care since these are synthetic benchmarks and therefore not always 100% correct!
Then I timed how long it took before 1 685 mb big file and 1 6419files/455 mb big folder was written from my primary disk to the test disk and afterwards being returned to the primary disk.

Finally, and this is very subjective: how loud are the drives? I don't have a db-meter so I just listened very carefully and quoted the disks on a scale from 1 to 10, 1 being very loud and 10 being almost silent

Ok, so far for all the jib-jab theory, let's take a look at the numbers...

HD Tach benches

Straight from their website since I found it difficult to explain what it exactly does:

"HD Tach is a physical performance hard drive test for Windows 95/98/ME and Windows NT/2000. In Windows 9X/ME it uses a special kernel mode VXD to get maximum accuracy by bypassing the file system. A similar mechanism is used in Windows NT/2000.

The HD Tach sequential read test is a little bit different from other benchmarks. Most benchmarks create a file on the hard drive and test within that file. The problem is that modern hard drives use a zone bit recording technique that allows different read speeds depending on where the data is located. Data on the outside of the drive is much faster than data recorded on the inside.

HD Tach reads from areas all over the hard drive and reports an average speed. It also logs the read speeds to a text file that you can load into a spreadsheet and graph to visually read the results of the test.

In addition to sequential read, HD Tach tests the drive's random access time. Random access is the true measure of seek speed. Many drives advertise sub 10 millisecond seek speeds, but seek speeds are misleading. Access time is the time it really takes to read data, not just the time it takes to move the head to the proper cylinder. To calculate access time the software must read a single sector off of the hard drive. By reading a sector the drive can not respond to the command until the sector is available, so rotational latency + seek time = access time.

A 7200 rpm drive has a rotational latency of 4.15 ms on average. A 4500 rpm drive has a rotational latency of 6.67 ms on average. So a 7200 rpm drive with a seek of 12.5ms has an access time that is just as fast as a 4500 rpm drive with a seek of 10.0 ms.

Finally, HD Tach tests the drive burst speed. The burst speed is the speed that data can be accessed from the drive's on-board read-ahead memory. This measures the speed of the drive and controller interface. Currently UltraDMA/66 IDE drives can reach 66mb/s. Ultra2LVD SCSI drives burst up to 80mb/s. "

And as I suspected, couldn't have it explained better myself ;)

In the next graph I mention the random access time since this is what interest me most



An estonishing result: the regular DiamondMax (which was only used as a reference drive) takes a headstart while the two 8 mb buffer disks share second place, the WD 800BB takes third place and the Seagate is performing quite badly

Quite a big difference between what you expect to get and what you really get no?

Now lets take a look on how the write & read tests go:



The average write speed of the 2 Maxtor drives is far beyond that of the other drives, the WD 800BB also seem to outperform the JB version in this test. The read test however is pretty even among the competitors with the Max Plus 9 in the lead.

The screenshots of test results can be found below:

Maxtor DiamondMax
Maxtor DiamondMax Plus9
Seagate Barracuda IV
WD Caviar 800BB
WD Caviar 800JB

SiSoft Sandra benches

The full name of the version I've been using here is "SiSoft Sandra 2002 professional sp1" . Sandra, who is Sandra and what does she do? Well, she is the System ANalyser, Diagnostic and Reporting Assistant. This program is capable of testing and benching about every piece of hardware that's inside your personal computer. The results then can be compared with other typical systems.

For this roundup the only section I used is the file system benchmark. It performs several read and write tests (buffered,random and sequential) and a seek test. These tests are then mixed together with a drive index number as a result, the higher this number is, the better...





Here the Maxtor DiamondMax Plus9 takes a giant leap compared to the others, while the Seagate is making up for the bad result in the previous test. Both the WD's seem to have problems getting a high score with this benchmark. Our reference drive is amazingly still going strong!!

Now onto what really matters, how fast can they read/write files when you starting copying files around...

Writing/reading files from disk to disk

For this benchmark, I created two folders on my primary hdd, one containing a single file of 685Mb big, the other 6419 files good for 455Mb. With the good old stopwatch in one hand and a finger of my other hand on the left mouse button I started copying folders from the primary disk to the test disk and vica versa. Now we will see if 8Mb buffer hdd's are really worth the extra $/?...





Poor Seagate, the numbers tell me you need to retire (for now)! The other 2Mb cache drives are holding up quite well and finally, the 8Mb buffer drives seem to be of some use! Will they be able to make us drool in the last run?





Yes they do! Although the DiamondMax Plus9 is quite a bit slower on the writing part, the two 8Mb buffer proof their worth every penny. It seems this last run was the hardest one, concidering the fallback for the regular Diamond Max while the good ol' Barracuda IV is finally able keeping up with that other 2Mb cache drive, the WD 800BB.

Soundcheck

This is more a personal impression then a realistic test since, as I mentioned at the beginning of this roundup, I don't currently have a sound level capturing device (dBa-meter)

From my own point of view, the Seagate Barracuda IV is one of the most silent hdd's for sale at the moment but what did I think about the other drives in this roundup?





Noiselevel wise, the Seagate still rulez, closely followed by the Maxtor DiamondMax Plus9 (the Fluid Dynamic Bearing really make noise disappear) and the WD800JB. The Maxtor DiamondMax takes fourth place and the WD800BB comes last here.


Conclusion

The Maxtor DiamondMax Plus9 is THE device for the moment closely followed by the WD 800JB (both the 8Mb buffer drives) These are the drives you want to be looking for when you want performance and what's more their pretty silent operation comes for free.

When you are really after silence and don't need that last drop of performance you should buy the Seagate, it is fast enough for daily usage.

The two other testsubjects aren't bad either, they are good mid-end products for a reasonable price.

One last thing I forgot to mention is the warranty period for the drives: the two Maxtors and the WD800BB carry a 1 year limited warranty, the WD800JB and the Seagate seem to be less reliable (as in "they escaped the new cost reducing policy by the hdd manufacturers ;)" ) and still carry a limited three year warranty!!

Well, that's all folks! Hope you enjoyed reading this and find the results usefull if you're considering buying a new hard disk drive at this time. With a 10.000rpm IDE drive already announced by WD and the SATA drives around the corner, waiting some time before buying a new storage device might be a good idea :)

Specials thanks go to Wim from Loveno Webstore for kindly supplying the test HDD's used in this roundup.

If you have any questions or remarks you can mark them @this thread on our forums.

//biCker