Is the Maxtor DiamondMax 80 as fast as the IBM Deskstar 75GXP?

 Small Dog has refurbished G4 towers! has refurbished towers, laptops, displays, etc., with 12 month warranty.

Original Posting 11/24/2000; Re-test 11/25/2000
© 2000 Rob Art Morgan, publisher of BARE FEATS

Re-test? Why? Within minutes after posting the first test results for the Maxtor 80G, I received an email from a reader who had purchased the drive just before reading my report. (I got sustained write speeds of only 12MB/sec.) He was sick. When he tested it himself, he was sicker. He dug around on the Maxtor tech support articles on their site and came up with a tech note explaining the "molasses" mode on the DiamondMax drive. Turns out that the drive is set at the factory to do "write and then read" verification the first 10 times you cycle power on the drive. After the reader cycled the power 10 times on his drive, his write times jumped from 12MB/sec to 29MB/sec. I tried the same recipe on my test drive and re-tested. Big difference! (Thanks, CW! Isn't Internet a hoot?)

When I get my hands on a new drive, the first test I run to get a flavor of potential speed is ATTO's ExpressPRO-Tools.

I've started using Intech's SpeedBench 1.5 to gauge the random read/write speed. They test at various block sizes. I chose 1024K.

I have two tests.
1. I duplicate 3 copies of the 27MB Photoshop document at the same time.
2. I duplicate a 101MB folder with 265 files, most of them small.

Duplicating files on the test drive forces it to read and write to itself simultaneously. The result is expressed in megabytes per second.

I play all frames of a 5.5 second, full screen, uncompressed QuickTime movie (176.7MB document). The results are expressed in MB/sec. If the drive kept up with "real time," it would clock 32MB/sec. Since neither drive achieves that, you would need to use some kind of compression scheme.

Consumers are excited about the
Maxtor DiamondMax 80G. All that storage on one drive is hard to resist. I say the Maxtor is a GREAT buy! It's not as fast as the hot IBM 75GXP but it's pretty fast. It can't be beat for the money. If you shop around, you can buy a Maxtor 80 for as little as $254. The 75GB 75GXP will cost you at least $363. Case closed.

Caveat: I tested the 45GB version of the IBM 75GXP. I would have preferred to have tested the 75GB version against the Maxtor 80 but I didn't have one. However, if you examine closely the article by Storage Review, you'll see they measured the speed of the 75GB and 45GB drives as virtually identical. So, no email yelling "recount," please.



The Maxtor DiamondMax 80GB was reviewed by Storage Review. They also reviewed the IBM 75GXP 75 GB. (Tests done using Windows machine.)

The Review Zone has a Clash of the Titans compares the two drives to each other (Tests done using Windows OS).

Tom's Hardware Guide compares the Maxtor 80GB to the 75GXP and drives from 4 other manufacturers (test done using Windows OS).



This is a hot topic among readers, especially those trying to use Ultra ATA drives with their "legacy" Power Mac to do Digital Video or Analog Video. Before you panic and return the drive, make sure you read Apple's Tech Note on the "16 Most Likely Causes of Frame Dropping." Of the 16, only 2 are drive related!



The Maxtor DiamondMax 80GB is available from MaxtorDirect for $310. Check also with,, Other World Computing, and for even better prices. The best price I found was at Upgrade Planet for $254.

The IBM Deskstar 75GXP 75GB drive is selling for $530 at IBM-Direct. Check with,, Other World Computing, and for better prices. The best price I found was at Upgrade Planet for $363.

If you need a PCI Ultra ATA controller, I recommend the UltraTek/66 is available from SmartDisk/VST Technology direct for a special price of $99. I think the UltraTek/66 is a great value because the price includes a copy of SoftRAID (worth $150) and dual cables. It's like you're buying SoftRAID and getting the card for FREE!! Use SoftRAID to create a striped array using two or more drives. Check also with,, Other World Computing, and for even better prices.



The TEST machine: An Apple G4/400 Sawtooth and 512MB of "222" PC-100 memory. Mac OS 9.04, VM off, ATALK off, clock display off, minimal extensions.

The Ultra ATA drives were:
IBM Deskstar 75GXP 45GB 7200 RPM Ultra ATA/100 with 2MB cache.
Maxtor DiamondMax 80GB 5400 RPM Ultra ATA/100 with 2MB cache.

The sustained read/write benchmark was run using
Express-Pro Tools 2.3.2 benchmark test with 8MB maximum file size and system disk cache disabled. Sustained Rate is displayed in the charts. Peak rates can be impressive but do not reflect typical drive performance.

Intech's Hard Disk Speed Tools 3.1.1 include SpeedBench 1.5. It's a great gauge of the random read/write speed. Although you can choose various block sizes from 1K to 1024K, I chose to display the 1024K results.

For the DUPLICATE THREE BIG DOCS test, I copied a 27MB
Adobe Photoshop document onto each drive and made three copies. Then I timed how long it took to duplicate the THREE copies of the photo (81MB total) on the same drive. That's a simultaneous read/write of 3 large files.

For the DUPLICATE 265 DOCUMENTS, I copied the Unreal Tournament 425a folder (101MB) onto the drive and duplicated it. The "megabytes per second" was calculated using the formula M=size/time * 2.

Thanks to a reader named Alex, I was able to use a 5.5 second segment (176.7MB) of a full screen, uncompressed QuickTime movie he created. I set it to play all frames as fast as possible. Using a stopwatch, the playback is timed to the nearest tenth of a second. Since I know the exact size of the document, I calculated the "megs per second" playback rate. All frames at "Real Time" would be 32MB/sec.

GO TO bare feats HOME PAGE

© 2000 Rob Art Morgan, publisher of BARE FEATS
Gotta Question? Comment? Want to advertise?