Click for Data Doubler kits!
Click for Data Doubler kits!


A click on the banner shows your site support to my sponsors

AccelerateYourMac.com logo
Maxtor's  FAST 40GB DiamondMax Plus
40GB 7200 RPM/2MB Cache ATA/66 IDE Hard Drive
By Mike
Published: 2/21/2000
(Updated 2/23/2000 for U2 Cheetah Comparison)
Introduction:

Update: The drive reviewed here was a January 2000 production model ATA/66 drive. If you're seeing low write speeds with the latest ATA/100 model Maxtor drives, try testing again after about a dozen power cycles (shutting down/power off, then starting up again). Maxtor ATA/100 series drives have a write verify feature that is reportedly disabled after 10 power cycles. Thanks to Mike Felsmann for reporting this. See this FAQ IDE Hard Drive Topic for more details, a list of drives affected and links to Maxtor's tech note.

Maxtor's latest, largest and fastest drive to date is the DiamondMax Plus 40. The drive features an ATA/66 interface (backward compatible with most older IDE controllers) as well as a 2MB cache and 7200 RPM spindle speed. With the highest density of any Maxtor drive to date (perhaps any IDE drive) I expected record-breaking performance and I was not disappointed.

As shown in the graphs of test results below, comparing the 40GB drive (Plus 40 series) to Maxtor's 27GB 6800 series (also 7200 RPM/2MB cache) and the OEM B&W G3's Quantum 6GB EX drive showed impressive gains in most tests. Duplicating a 400+MB folder in the Finder showed surprising performance advantages over even the 27GB model - appx. 40% faster. It was almost twice as fast at this test as the stock OEM 6GB Quantum drive. Peak rates were limited by the B&W G3's onboard ATA/33 controller, but sustained rates were the highest I've seen on any drive to date - near 30MB/sec according to ATTO's latest benchmark. (Full test results are below).

[Update: An Ultra2 Cheetah SCSI drive owner questioned how this Maxtor Plus40 compared to a Ultra2 PCI Card/Ultra2 10,000 RPM drive. Since I have an ATTO U2 SCSI card and 2nd generation U2 10,000 RPM Seagate Cheetah 4.5GB drive, I've added the results of tests with that card/drive in the same B&W G3 system to the results below. I'm sure there are faster U2/10K RPM drives than the Cheetah I have, but currently that's the only 10K RPM U2 drive I have available. To avoid any questions on quality of cables or termination, I used a Granite Digital cable and addon 'forced perfect' terminator that together cost almost as much as the Maxtor 40GB drive. (About $170 for the pair based on last year's prices.). The ATTO's U2 PCI Card LEDs showed the interface was running in Ultra2 mode.]

Pricing and Availability:
As noted in the Sunday (2/19/2000) news - CompUSA had this drive on sale for $249.99. Best Buy and Staples regular price at this date was $299 and change. OEM versions (drive only/no retail package) may be less but as of this date I've not seen the OEM version selling for less than the CompUSA retail box sale price. [BTW: I paid $299.99 for this drive at Best Buy on Saturday - after talking to the manager I was given a credit to match the CompUSA $249.99 price. All Maxtor drives I have tested at this site have been purchased at retail outlets (no mfr loaners, etc. all retail store/retail box purchases).]


What's Included in the Maxtor Retail Box:

  • 40GB DiamondMax Plus drive
  • Installation guide (geared at PCs) with jumper settings
  • Maxtor Installation floppy disk (for PCs - not needed for Macs)
  • ATA/66 dual drive cable
  • Metal 3.5"->5.25" mounting bracket
    (may be useful for some Mac clones - not needed for Apple Macs)
  • Adobe Photoshop 5LE CD (dual platform - may not be in all packages)

DiamondMax™ Plus 40 GB Drive Specs:

  • 7200 RPM spindle speed
  • 2MB cache
  • <9ms Avg. seek time (<20ms Max, 1ms track to track)
  • Avg.. latency 4.17ms
  • Start time (0-drive ready) <10sec typical
  • Up to 43.2MB MB/sec transfers to/from platter,
    Up to 66.7MB/sec from interface (with ATA/66 controller)
  • Component design life: 5 years (3 year warranty)
  • Start/Stop cycles: 50,000 minimum
  • Annual return rate: <1%
  • Drive height: 1.02 inches
  • 8 Heads/4 platters

For a complete list of drive specifications, see Maxtor's PDF file.

Tests in ATTO Tools showed the drive was capable of 28-29.4 MB/sec Read/Write sustained rates (tested on B&W G3's ATA/33 onboard IDE controller).

About IDE Slave Support: For those that are not aware, a second IDE drive on the same cable is called a Slave (the primary drive is the Master). Mac models before the Beige G3 did not support slave drives. In addition, Rev 1 Beige G3 and rev 1 B&W G3s do not support IDE slave drives (although I was able to use a Maxtor 10GB IDE drive as a slave in my rev 1 B&W G3). See my Beige G3 IDE Upgrade and B&W G3 IDE Upgrade articles for more info. This drive should also work well in a Beige G3 (see my Beige G3 IDE drive upgrade article).


Observations:
The Maxtor 40GB DiamondMax Plus drive seems very quiet. Initial observations shows the drive runs a bit warmer than some IDE drives, but temperature was not a problem in operation. The drive is rated for up to 55°C operation.

There were no problems seen with the drive set as a slave with the original 6GB Quantum EX drive as a master in the B&W G3 rev 2 system. (Rev 2 B&W G3s have a revised IDE controller that addresses the issue of data corruption seen with rev 1 B&W G3s. (See my B&W G3 rev 2 page for pictures/info on how to determine your system revision.)

Test System Hardware Summary

  • Apple B&W G3 rev 2
  • 400MHz CPU Clock speed
  • 256MB SDRAM (322 clock dimms)
  • Stock 6GB IDE hard drive (Quantum EX 5400 rpm, 512KB cache)
  • DVD drive (OEM)
  • ATI Rage128 rev2 graphics card in 66MHz PCI Slot
  • OS 9.0, VM off, QT 4.0.3 Pro, QD3D 1.6., OpenGL 1.1.2, standard extensions set (including Multiple Users, PC exchange, etc.)
  • Memory Control Panel disk cache was set to automatic (8160KB cache size with the amount of installed RAM in the system)

Disk Driver Used: Apple's Drive Setup v1.8.1 (OS 9 release) was used on all drives in this article.


How I Tested:

Since the performance of a hard drive is always slower as the drive fills with data (the outer tracks are the fastest), each of the 3 drives (the Maxtor 40GB, Maxtor 27GB and original Quantum EX 6GB) was erased in the Finder (Mac OS Extended format) before each test. This allowed an accurate indication of the maximum performance of each of the drives. In addition the system was rebooted before each test was run to ensure as clean a RAM/Cache/System state as possible.

The following tests were performed:

  • Duplicate a 414MB Folder with 256 Files
  • Time to Boot OS 9 (appx. 4GB of files on the disk, OS 9 full extension set including the software for a G4 CPU upgrade)
  • ATTO Tools 2.3 Benchmark (0-8MB test setting)
  • MacBench 5.0 complete Disk Test Suite


Real World Test Results:

Duplicate 412MB Folder times

[The 10K RPM U2 Cheetah 4.5GB drive took 52.34 seconds to duplicate the folder.]

Note the dramatic speed increase the 40GB drive provides, even over the 27GB ATA/66 drive (same RPM and cache size). The 40GB DiamondMax Plus drive was approximately 40% faster than the 27GB drive and 98% faster (almost twice as fast) as the OEM 6GB drive.

Time to boot OS 9

[The 10K RPM U2 Cheetah 4.5GB drive took 95.1 seconds to boot.
See notes below on slower IDE boot times w/U2 PCI card installed as well
]

Two notes on the OS 9 boot time. The B&W G3 had a CPU upgrade installed and therefore had to load both an extension and control panel. Also the OS 9 'delay' during initial mounting of volumes in the Finder was present. OS 9 appears to be searching for network volumes at first boot (I think there was some fix/workaround for this) despite the fact the test system was not connected to a network.

I was surprised at the times for the U2 Cheetah boot tests. I can only guess that booting from a PCI card connected drive versus the onboard IDE adds additional delays. The ATTO U2 PCI SCSI card was installed in slot 1 - the highest priority PCI slot in the B&W G3. In fact, with the PCI U2 SCSI card installed - booting from the IDE drives took longer (Maxtor 27GB drive took 90 seconds with the U2 SCSI card/U2 Cheetah installed).


Since the U2 Cheetah owner mentioned how much faster a 10K cheetah/U2 SCSI drive would be than IDE drives to launch apps, open files etc., I also tested times to launch Photoshop 5.5 (these are stopwatch times, so consider those within 1/2 second or so literally a wash due to the human error factor). Since I had removed the 40GB Plus drive the tests below show the lower performance Maxtor 27GB drive for comparison to the U2 Cheetah. The Plus40 would likely be faster than the older 6800 series 27GB Maxtor of course.

  • Ultra2 Cheetah 4.5GB: 6.60 seconds
  • Maxtor 27GB IDE: 5.71 seconds

Load 10.1 MB Image file:

  • Ultra2 Cheetah 4.5GB: 1.95 seconds
  • Maxtor 27GB IDE: 1.75 seconds

Save 10.1 MB Image file:

  • Ultra2 Cheetah 4.5GB: 1.72 seconds
  • Maxtor 27GB IDE: 1.28 seconds

As the above shows, don't expect a U2 single drive to show any real benefits to the average user over the best IDE drives on the market currently. If you need massive RAID, etc. SCSI may be worth the extra money to you, but multi-drive RAID is very expensive and not needed by 99% of end users. The value of these new generation IDE drives is clear. (The test system has an ATA/33 controller. Performance would be less in a Beige G3 with older IDE controller, perhaps better in a G4 with ATA/66 interface.)


Benchmark Test Results:

ATTO Benchmark Results:

To make it easier to see comparisons of the stock drive, Maxtor 40GB and Maxtor 27GB drives, I graphed the results of Peak and Sustained rates. The actual ATTO Tools screenshot is shown farther down. The ATTO graph of performance across file size can be interesting to see. (You can often detect the cache size of the drive by the graph for instance.)

Sustained Rates (B&W G3 ATA/33 interface)

[The 10K RPM U2 Cheetah 4.5GB drive Sustained rates were Read 18.74/Write 18.81 MB/sec.]

From comparing the sustained write rates of the 40GB vs 27GB, you'd think write caching was not enabled on the 27GB drive (same Apple OS 9 driver was used). However a peek at the actual graph of data rate below shows there is a peak in the write speed that would indicate the drive's cache is being used. The peak rate graph below also shows this. The 40GB drive's higher platter data density must be at least partly responsible for its better sustained write rates.

Peak Rates (B&W G3 ATA/33 interface)

[The 10K RPM U2 Cheetah 4.5GB drive Peak rates were Read 18.88/Write 23.47 MB/sec.]

I'm sure peak rates of the ATA/66 IDE drives were limited by the B&W G3's ATA/33 (Max 33MB/sec) onboard IDE controller. I've seen over 50MB/sec rates in the G4 system which has onboard ATA/66. (Tests in the Apple G4 system will be run to see what the actual rates are with the new Maxtor 40GB drive.)


Here's the actual ATTO Tools screens used for the above graphs. (By observing the spikes in the graph you can literally see the size of the drive's cache in many cases.)

Atto results - OEM 6GB drive

Atto results - Maxtor 40GB drive

Atto results - Maxtor 27GB drive

[Here's the graph of the U2 10K Cheetah 4.5GB drive:]

Atto results - U2 Cheetah 4.5GB


MacBench 5.0 Results:

The image below shows the MacBench 5.0 Disk Tests Suite results from the B&W G3. All drives were empty (no files present) and formatted just before the test. As with all tests, a clean reboot was performed before executing the benchmark.

Complete MacBench 5.0 Disk Test Suite

[Here's a comparison of the U2 10K Cheetah vs. the Plus40 IDE drive in the B&W G3:

Random R/W 512K scores were faster with the U2 drive, other scores lower. Perhaps the fact the IDE volume is about 9x larger than the U2 drive affected random access scores but partitioning the IDE drive into a 4.5GB volume would have shortened the full stroke of the volume (compared to the full drive stroke using 1 partition the full size of the drive). This shorter stroke would have been unfair in my opinion so I left the volumes at their full disk sizes. ]

U2 Cheetah 10,000 RPM vs Plus40 IDE drive


Audio Professional's Comments on the Drive: Here's a post on the Plus40 drive's performacne from my forums from an audio professional (using onboard IDE, not a PCI IDE card):

" ...10k Cheetahs and Atlas are a real pain in the recording studio because of the high-pitched whine that interferes with proper mixing. Really, I could care less about the Maxtor's 40GB size - audio drives rarely need to be larger than 18GB, at the most. Anything beyond that is luxury. What impresses me about the Maxtor is what I've said below: no errors, quiet operation, and massive throughput. Last night I put both the PT session AND the DV movie files and played them back together, and had no problems. Try that with a PCI-based SCSI disk.

And about reliability and craftsmanship - all drives fail; it's inevitable. I've watched helplessly as Cheetahs spin down in the middle of a recording session with clients hovering over my shoulder with that sort of blank look on their faces. Internal or external, no drive is completely reliable.

As I said below, the only reason that many in my profession continue to buy SCSI disks is interstudio compatability, i.e. taking your disk to work in different studios, etc. SCSI disks are fine for servers and ultra high-end Avid video, but for everyone else...
DJS
"


How Does the Plus40 Perform in a Firewire/IDE Case? Good Question. Click here for the answers including a reader's report of Final Cut Pro use with the Plus40 in a Firewire/IDE external case.


When time allows, I'll be updating this page with tests in an Apple G4 system with onboard ATA/66 controller. A page on installing the drive and formatting it with Apple's drive setup is also planned.


Plus40 Tests in Beige G3: Chad Mathes sent test results of the Plus40 drive in his Beige G3/333, including comparisons to the original 9.1GB UW SCSI drive (his Beige came with the UW SCSI card and drive option).



Related Links:

Back to XLR8YOURMAC.COM

Copyright © Mike, 2000. All rights reserved.
All brand or product names mentioned here are properties of their respective companies.

Users of this web site must read and are bound by the terms and conditions of use.