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

Quantum's Atlas 10k II vs. Seagate's Cheetah 36LP: A Grudge Match!
  July 16, 2000 Author: Eugene Ra  
Atlas 10k II evaluation unit provided by Quantum Corporation
Cheetah 36LP evaluation unit provided by Seagate Technology
See also Quantum Atlas 10k II review
See also Seagate Cheetah 36LP review
See also One Last Time: The Quantum Atlas 10k II

Introduction

Towards the beginning of the year (February 2000), StorageReview.com was proud to bring readers the world's first reviews on two highly anticipated drives: the Quantum Atlas 10k II and the Seagate Cheetah 36LP. An interesting contrast exists in the time it took these two units to reach the general market. Seagate, for its part, managed to fill channels with the 36LP in less than a month after our review. Quantum, on the other hand, experienced considerable delay. On June 1st, the company issued a press release announcing "volume shipment" of the drive. Even so, as evidenced by threads in the SR Discussion Forum, eager potential purchasers can't find the drive for sale at any of their favorite retailers. Hopefully we'll see it soon.

A couple months ago, a representative from Seagate remarked that the Cheetah 36LP figures in our database looked a bit low when compared to its smaller brother, the Cheetah 18XL. Pre-release firmware was the suspect according to the rep, 36LP's possessing "release" firmware should be the fastest of the 4th generation Seagate 10k units. This claim perplexed me a bit. As outlined in our 18XL review, Seagate took great pains to explain that the 18XL's physically lighter actuator should result in faster-than-36LP operation. Nevertheless, we took Seagate up on the offer for a retest- SR readers should always have access to the most accurate and most current performance figures around.

As we witnessed the increasing delay in the Atlas 10k II's general availability, it became clear that we should also ask Quantum if they'd like to ship SR a unit with revised firmware for additional evaluation. After all, our initial review (preview??) hit the street over three months before the firm's announcement of volume shipments. Quantum complied by sending us another 10k II.

In this follow-up article, we're going to take a look at how each revised drive compares to its pre-release form. And, for good measure, we'll stack them up head-to-head to see which comes out on top . Both units feature 10,000 RPM spindle speeds, representing the pinnacle of nearly four years of development. When it comes to claimed specifications, however, Quantum's drive impresses. The Atlas specs out with a 4.7 millisecond seek time compared to the 5.2 milliseconds claimed by the Cheetah. Quantum crams 36 gigs of data onto five 7.3 GB platters rather than Seagate's six 6.1 GB disks. Finally, the 10k II features hefty 8 meg buffer; Seagate's unit provides just 4 megs. Both units, of course, are backed by a five year warranty.

Our original Cheetah 36LP possessed firmware revision 0001. The unit tested here features revision 0003. Similarly, the original Atlas 10k II's firmware was D620 vs. the newer unit's DA40. Let's turn to the figures and examine exactly what revised firmware brings to the table.

WB99/Win2k Low-Level Measurements

 Testbed II Low-Level Measurements Details... 
Windows 2000 Professional using NTFS
Quantum Atlas 10k II 1st Sample - 7.8|
Quantum Atlas 10k II 2nd Sample - 7.9|
Seagate Cheetah 36LP 1st Sample - 9.0|
Seagate Cheetah 36LP 2nd Sample - 9.0|
Windows 2000 Professional using NTFS
Quantum Atlas 10k II 2nd Sample - 41467|
Quantum Atlas 10k II 1st Sample - 40167|
Seagate Cheetah 36LP 2nd Sample - 36167|
Seagate Cheetah 36LP 1st Sample - 36133|
Windows 2000 Professional using NTFS
Quantum Atlas 10k II 1st Sample - 25000|
Quantum Atlas 10k II 2nd Sample - 25000|
Seagate Cheetah 36LP 1st Sample - 24100|
Seagate Cheetah 36LP 2nd Sample - 24100|

Click here to examine STR graph for these drives

The access times exhibited by both revised drives fell in line with the measurements gleaned from the original units. The Cheetah 36LP's seek times hovered right around 9.0 milliseconds. Though the seek time in the 2nd sample of our Atlas 10k II reads 7.9 milliseconds vs the 7.8 turned in by the original, the difference is a cosmetic one presented by the peculiarities of rounding out to just two significant digits. The measured times hovered around 7.85 milliseconds, with the original Atlas' figure coming in just a hair below and thus winning itself the insignificant 7.8ms figure. In any case, however, the Altas 10k II's access time remains extremely impressive, clocking in at over a full millisecond below that of the Cheetah 36LP.

Due to its increased areal density, the Quantum sports an advantage over the Seagate when it comes to sequential transfer rates. Don't be fooled by the Atlas' impressive 40 MB+ figure... the drive maintains this speed only in its relatively small outer zones. The Seagate, on the other hand, features a more "graceful" decay. Though the Atlas sports a significant outer zone lead, the difference becomes negligible as inner zones are reached. Please examine the STR graphs presented above for more detail.

WB99/Win2k WinMarks

 Testbed II  WB99/Win2k WinMarks  Details... 
Windows 2000 Professional using NTFS
Quantum Atlas 10k II 2nd Sample - 8763 |
Quantum Atlas 10k II 1st Sample - 8247 |
Seagate Cheetah 36LP 2nd Sample - 7040 |
Seagate Cheetah 36LP 1st Sample - 6170 |
Windows 2000 Professional using NTFS
Quantum Atlas 10k II 1st Sample - 24633 |
Quantum Atlas 10k II 2nd Sample - 23033 |
Seagate Cheetah 36LP 2nd Sample - 19667 |
Seagate Cheetah 36LP 1st Sample - 17167 |

When it comes to the Business Disk WinMark 99, the Atlas 10k II manages to improve on its already amazing score. The second sample's 8.7 MB/sec figure tops the original's by a margin of over 6%. In an interesting contrast, however, the newer Atlas actually regresses behind the older in the High-End test, trailing by 6%.

The newer Cheetah 36LP manages substantial improvements in both the Business and High-End test. Its scores of 7.0 MB/sec and 19.7 MB/sec respectively in each test are nearly 15% improvements over our first sample.

When the Quantum and Seagate are pitted against each other in WinBench99, the Quantum prevails. The Atlas leads the Cheetah by a formidable 25% in the Business Disk WinMark. Quantum's drive also reigns in the High-End WinMark where it best the Seagate by over 17%.

So far, things look quite rosy for the Atlas 10k II. Before declaring it the victor, however, we should examine what changes await us in the far more significant IOMeter.

IOMeter Performance

 Testbed II  IOMeter - Workstation Access Pattern - Total I/Os per second Details... 
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|

The newer Atlas 10k II does not fare as well as our original sample in IOMeter. It starts with the new unit lagging the old by about 2% in a Linear Workstation scenario... this gap widens to 5% and 6% respectively under Moderate and Heavy loads.

Seagate's drive, on the other hand, exhibits some slight improvements across the board. The new drive improves over the original by a tiny 1% in the Linear Workstation load. This improvement, however, peaks to a 6% gain under Heavy loads.

These reversals lead to an interesting change in the Cheetah vs. Atlas story as told by IOMeter. Judged with our original units, the Atlas 10k II possessed a marginal advantage over the Cheetah 36LP. Now things are reversed. Due to its superior seek time, the Atlas retains a lead of 4% under a Linear load. In every other case, however, the Cheetah leads. Very Light, Light, and Moderate loads peg the Cheetah 5% ahead of the Quantum. The margin increases to a significant 10% under a heavy load.

Conclusion

Noise and operating temperatures remain unchanged between old and new versions of both drives. In other words, both are a bit noisier and louder than today's cool, whisper-quiet ATA drives. As is always the case with 10k drives, care should be taken to preserve a well-ventilated, cool case environment.

Overall, what we've witnessed here are small changes between firmware revisions of two drives that add up when they're compared against each other. Though the Atlas 10k II improves upon its already phenomenal Business Disk WinMark score, its regression in IOMeter combined with the improvements delivered by the Cheetah (not to mention the current price gap... hopefully the Atlas will come down) drive us to give the nod to the 36LP. Seagate's latest would thus garner our Leaderboard spot if it weren't for their own Cheetah X15 drive . All is not over yet, however. Fujitsu's impressively-spec'ed MAJ3182LP waits in the wings. Though we've been disappointed in the past with Fujitsu drives not living up to the performance that specs hint at, we nonetheless have high hopes for the MAJ. Expect a review within a week!

Quantum Atlas 10k II MSRP (as provided by Quantum): $1106
Quantum Atlas 10k II Specifications
Seagate Cheetah 36LP MSRP (as provided by Seagate): $969
Seagate Cheetah 36LP Specifications
Initial Quantum Atlas 10k II review
Initial Seagate Cheetah 36LP review

HOME | ARTICLES | LEADERBOARD | PERFORMANCE DATABASE | REFERENCE GUIDE
COMMUNITY | RELIABILITY SURVEY | SUPPORT SR! | ABOUT SR |

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