|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
By Mike Breeden Published: February 1998 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Results with the ATTO benchmark - with Conley and FWB drivers | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
The ATTO Tools utility has a benchmarking function, which I personally like as it's fast and does not use the system disk cache (although it has an option to do so). Bypassing the system disk cache gives a better indication of the card and drive performance, eliminating the effect of a large cache setting in the memory control panel - which inflates scores in other benchmarks like MacBench 4.0. When the tests are run with the system cache enabled, peak reads and writes often reach 60-100+MB/sec (esp. with the PowerForce and other G3 cards with large fast backside caches). You can argue that this is the state you will using the system in, but for an accurate test of card and drive/driver performance, we want to remove as many other factors as possible. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
For this test all cards were installed in slot 1A (considered best by Apple, Adaptec and Media 100). In the future
I plan to show results of tests in other slots of the Tsunami motherboard in the PowerTower Pro.
The FWB driver was optimized using the standard control/optimize function.
The Conley driver was set to the default 256 I/Os setting, with fast read, fast write and fast Photoshop settings enabled. The Conley driver is primarily for Raid, but I was curious to see how it would perform in single drive configuration. I've asked Conley what effect (if any) the Fast Photoshop setting could have on these tests (non-Raid) - it may have affected results in the smaller sample size. The 32K scores below would be indicative of performance in a database application, while the larger sizes would be more applicable to typical file I/O and larger file applications like video capture. Most drivers, especially Raid, tend to be optimized for the larger file size I/O, but this is usually selectable in the array setup phase. Note: All rates are in MegaBytes/Second. Drivers: Conley denotes SoftRaid 2.0, FWB denotes Hard Disk ToolKit 2.06 (optimized). PR = Peak Read, SR = Sustained Read, PW = Peak Write, SW = Sustained Write.
Although the Stock SCSI and IBM AV 7200 rpm drive performs well, the 10MB/sec max motherboard scsi is far too limiting for this project. Remember that this test did not use the system disk cache, to better indicate the actual drive and scsi interface performance. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
TOP PERFORMERS: JackHammer Card and Conley Driver. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Some call the ATTO test a little too simple, so I also tested the cards using FWB's Hard Disk ToolKit BenchTest, and the scores are shown on the next page. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
- or - | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright ©
Mike, 1998. No part of this sites content or images are to be reproduced or distributed in any form without written permission. Please ask before using quotes or images. Users of this site must read and are bound by the terms/conditions of use | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||