Original Link: https://www.anandtech.com/show/41



One of the major issues facing a prospective hard drive purchaser is the decision between an ATA and SCSI hard drive. It often seems that ATA offers the tantalizing benefit of many gigabytes for cheap while SCSI promises better performance at a higher cost. SCSI drives do not benefit from the economies of scale that ATA drives enjoy; as the "performance" rather than "mainstream" offering, SCSI drives outright cost more to manufacture in addition to sporting beefier margins for distributors and vendors. The addition of a SCSI controller to a system, whether an add-on card or built in to the motherboard, can not be ignored. Finally, SCSI drives tend to incorporate later technologies and better specs: higher rotation speeds, larger buffers, faster seek times, etc. All this translates into lots of expense.

Once in a while a manufacturer will produce the same mechanical drive in both an ATA and SCSI configuration. Quantum has done so with its Fireball series, the latest "SE" iteration being available in Ultra ATA and Ultra SCSI models. We decided to pick up 6.4 gigabyte versions of both and compare them head-to-head. The Ultra ATA protocol boasts a maximum transfer rate of 33.3 MB/sec compared to Ultra SCSI's 20 MB/sec. In the tests we performed, the testbed's boot drive (Seagate ST34555W) was inactive while the SCSI Fireball SE ran the tests. In the ATA Fireball's case, only one ATA device can be active at any given time per ATA channel. Either way, the maximum transfer rates were moot… neither drive comes close to saturating their respective interfaces.

The Quantum Fireball QM36400SE has a rotational speed of 5400 RPM, the maximum currently found in shipping ATA drives. Its buffer is on the low side, a relatively skimpy 128k. Quantum rates the Fireball SE's average seek time at 9.5ms. The drives use 2.1 GB platters. All this combined to give the ATA version a very respectable showing in the 6.4 GB ATA roundup.

For this test, we used an ABIT LX6 motherboard (v1.1, bios v C7Q), a 266 MHz Intel Pentium II processor, a 64MB 10ns SDRAM DIMM, and a Matrox Millennium II PCI 4MB (bios vf1.2, PowerDesk v3.80). The tests were run at 1024x768 with 24 bit color at 85 Hz using small fonts. ZDBop's Startup Manager was used to prevent loading of background applications. ZDBop's WinBench 98 v1.0's Disk Test Suites were run on the two drives. The SCSI controller used with the QM36400SE-S was the Adaptec AHA-2940U2W. Read caching and write caching were enabled. The boot drive (Seagate Hawk 4XL ST34555W) contained Windows 95 OSR 2.1 patched with Intel's 82371xB INF Update. For the QM36400SE-A, Windows 95 OSR 2.1 was installed, patched with Intel's 82371xB INF Update and Microsoft's REMIDEUP.EXE fix. The boot drive (Western Digital Caviar AC31600) resided on the primary ATA controller with the QM36400SE-A on the secondary channel. Here is a link to the test bed information.



Ziff Davis Winbench 98

Quantum Fireball SE (UltraATA)

Quantum Fireball SE (SCSI-2940U2W)

Business Disk WinMark 98 (KB/s)

1254

1110

SS/Database

1128

987

WP

1522

1338

Publishing

1144

1046

Browsers

1426

1234

Task Switching

1822

1618

High-End Disk WinMark 98 (KB/s)

3722

3226

AVS/Express 3.1

2306

2008

Frontpage 97

2920

2818

MicroStation 95

7032

6090

Photoshop 4.0

3038

2488

Premiere 4.2

6952

5918

PV-Wave 6.1

2720

2264

Visual C++ 5.0

8220

7160

Disk/Read Random Access (ms)

16.3

18.0

Disk/Read Transfer Rate (KB/s)

Beginning

11800

11200

End

7170

6800

Disk/Read CPU Utilization (%)

5.9

3.9

Transfer Rate (KB/s)

11758

7388

The ATA Fireball SE bests the SCSI version by 13% in the Business Disk WinMark 98 and by 15% in the High-End Disk WinMark. Curiously the results for the ATA drive were also substantially better in the low-level tests in addition to the Disk WinMarks. Both drives supposedly have the same seek time and the same rotational speed (and thus latency) yet the ATA Fireball posted a 16.3ms average access time to the SCSI version's 18.0ms. The Disk/Read Transfer rates of the SCSI Fireball also lagged behind its ATA counterpart.

CPU utilization per megabyte transferred for both drives are virtually identical. There's a common perception floating around that ATA's CPU utilization even with busmaster drivers does not approach that of busmastering SCSI. CPU utilization scores of ATA drives when using the busmaster drivers including with Windows 95 OEM Service Release 2 indicate otherwise, however. SCSI does not have an advantage here.

ATA drives and controllers, on the other hand, do not offer a key SCSI feature: Command Queuing. This feature allows multiple requests to "queue" and remain pending on the SCSI subsystem as opposed to ATA's serial approach where one request must be finished before another is started. This parallel execution allows SCSI to reorder the requests into a sequence that minimizes head/actuator seek motions. Simply put, SCSI's advanced features should result in less degradation in multitasking and multithreaded environments.

We decided to turn to Adaptec's ThreadMark 2.0 Benchmark. Adaptec offers the benchmark as a more reliable indicator of overall disk performance in multitasking environments. ThreadMark is difficult to run consistently, however. To obtain repeatable results, the test machine's use of virtual memory must be disabled. Storage Review's testbed has 64 megs of RAM, enough to run most applications well, but -not- enough to guarantee stability without any virtual memory. With virtual memory enabled, ThreadMark results varied wildly, sometimes displaying sample deviations of over 100%. Disabling virtual memory contained the sample deviations to a much more usable 1%. The average of five trials is presented here.



Adaptec Threadmark

Quantum Fireball SE (UltraATA)

Quantum Fireball SE (SCSI-2940U2W)

Data Transfer Rate (MB/s - Higher is Better)

5.498

3.46

Average CPU Utilization (% - Lower is Better)

22.158%

13.214%

The ThreadMark results were even more surprising: the ATA Fireball clocked in at 59% faster! One would expect from SCSI's multitasking advantage that the SCSI Fireball would have at least closed the gap between itself and its ATA sibling. Strangely enough, it's the other way around. ThreadMark unleashed the ATA drive, making the performance gap even more glaring.

A possible explanation for the SCSI Fireball's inferior showing would be poorly optimized control logic on the drive. Quantum is not a newcomer to the SCSI arena, though; one would figure that the company could provide a decent SCSI implementation even at the Fireball's relatively low price point.

In conclusion, the ATA version of the Fireball SE clearly seems to be a better performer than its SCSI cousin. Its very difficult to draw further conclusions to apply to ATA vs SCSI in general, however. The only thing we can infer from the results is that the SCSI version of the Fireball SE is a poor-performing SCSI drive. Paying $150-$300 for a good Ultra SCSI controller and then pairing it up with a SCSI drive which costs about $100 more than its ATA counterpart while providing worse performance to boot doesn't make much sense. At the expense of 2 gigs, one could pick up the better performing 4.5 GB Quantum Viking for about the same money. Seagate's 7200 RPM Medalist Pro Series is the next drive that will be available in both configurations. We look forward to testing the ATA and SCSI versions and presenting you with the results!

Log in

Don't have an account? Sign up now