Reviews Leaderboard Database Reference Search StorageReview Discussion Reliability Survey Search About Contents

Fujitsu MAM3367
  October 1, 2001 Author: Eugene Ra  

Fujitsu MAM3367 Available Capacities *
Model Number
18 GB
36 GB
Estimated Flagship Price: TBD (36 GB)
Evaluation unit provided by Fujitsu Computer Products of America.


Seagate's revolutionary X15, the world's first drive featuring a 15,000 RPM spindle speed, stood alone for all of 2000 as the only drive to deliver next-generation performance levels. The drive proved popular both with enterprise/IT purchasers as well as individuals who simply had to have the latest in bleeding-edge performance. This year, in addition to the X15's successor, we've witnessed announced competition from IBM and Fujitsu. Even so, Seagate's X15-36LP is the only drive shipping to channels. IBM's drive, though announced in February and reviewed by SR in June, hasn't been spotted yet. Fujitsu's announcement came in May, so hopefully the MAM drive reviewed here will hit the channel soon.

Fujitsu's MAM series goes head-to-head with the Cheetah X15-36LP, directly challenging the hegemony carved out by Seagate's newest dynasty. In addition to its 15,000 RPM spindle speed, the MAM3367 features a specified seek time of 3.5 milliseconds, the lowest yet for a 36 GB 15k RPM drive. It packs 9 gigs per platter, matching the areal density of Seagate's drive and exceeding Big Blue's by 50%. Fujitsu equips the MAM with 8 megabytes of buffer, again matching Seagate while doubling IBM.

Available in 18- and 36-gigabyte capacities, the MAM squarely targets the highest of high-end applications, areas where speed is of utmost importance. Though this review sample features an Ultra160 interface, the family will migrate to Ultra320 as the improved SCSI interface starts to permeate the industry. Fibre Channel versions of the drive are also available. An enterprise-class 5-year warranty protects the drive.

SR normally reviews the 68-pin UW/U2W/Ultra160 versions of SCSI drives. At the time of testing, however, Fujitsu could only offer us an 80-pin SCA unit. To expedite results, we accepted the SCA drive. An adapter converted the drive down to standard 68-pin data and 4-pin Molex power connections. The results presented below also represent the 68-pin LVD unit.

Let's turn to some figures and see how the MAM measures up!

WB99/Win2k Low-Level Measurements

[an error occurred while processing the directive]

Click here to examine the STR graph for this drive

With a WinBench 99 measured access time of 6.0 milliseconds, the MAM virtually matches the X15-36LP. Subtracting 2.0 milliseconds to account for the rotational latency of a 15k RPM drive yields a measured seek time of 4.0 milliseconds, a bit off of Fujitsu's 3.5 ms claim.

Outer-zone transfer rates clock in at 57.2 MB/sec, a little bit short of the Cheetah X15-36LP's record-setting 60 MB/sec. The Fujitsu's inner-zone rates, on the other hand, come in at 44.8 MB/sec, virtually the same as Seagate's drive. The overall effect is a, relatively speaking, ruler-straight STR graph.

WB99/Win2k WinMarks

[an error occurred while processing the directive]
The MAM trails the Cheetah X15-36LP when it comes to the WinBench 99 Disk WinMarks. The Fujitsu's Business score of 9.2 MB/sec, a score that would have been breathtaking just a few months ago, trails Seagate's drive by nearly 18%. A High-End score of 26.4 MB/sec creates a similar margin.

IOMeter Performance

[an error occurred while processing the directive]
It's in SR's IOMeter indices (normalized averages of Light, Medium, and Heavy loads) that the MAM pulls up to the X15-36LP. A File Server Index of 357.44 matches the score turned in by Seagate's drive. The MAM's Database index of 364.34 lags the Cheetah by an insignificant 1%. It's only in the Workstation Index that the gap widens, though at less than 3% it's almost negligible.


How about heat and noise? Fujitsu often excels in these areas. With the MAM, unfortunately, a different story arises. Seek noises are slightly louder than those heard from the X15-36LP. It's idle noise, however, where a true difference arises. The MAM exhibits a high-frequency pitch that we haven't heard since the days of second-generation 10k RPM drives. Though it's likely irrelevant in the server situations that the MAM targets, it may bother those who utilize the drive in a high-end, single-user system. Heat levels remain on par with the X15-36LP... meaning that though it's not as scalding as IBM's Ultrastar 36Z15, the drive gets hot to the touch outside a drive cooler in our smallish testbed case. Active cooling may be warranted in many situations.

Overall, while the MAM is a decent first effort (superior to the Ultrastar 36Z15 overall), it can't quite match the performance of the Cheetah X15-36LP. The latter provides equivalent if not superior IOMeter performance coupled with decidedly better WinBench 99 scores. To Fujitsu's credit, by the narrowest of margins, the MAM turns in the highest IOMeter file server index recorded to date. The MAM also, unlike the Ultrastar 36Z15, matches Seagate spec-for-spec... no compromises were made in areal density or buffer size. Hopefully Fujitsu's 15k RPM line will evolve into a competitor that gives Seagate a run for its money... the manufacturer has already proved that it's capable of such in the 10k arena.


Copyright © 1998-2005, Inc. All rights reserved.
Write: Webmaster