Click for Portable Storage Solutions!
A Click on this Banner shows your site support to my Sponsors
IBM's TravelStar 32GB, 5411 RPM drive
Performance Comparisons & Install
Since the PowerBook Firewire G3/500 test system already had an IBM 20GB Travelstar installed it was not possible for me personally to compare performance with the original Toshiba drive. However reader Charles Eicher sent results of his tests with a PB G3/500 and the original Toshiba 12GB drive. (My spring 2000 PowerBook Firewire review has results of tests with a 12GB OEM Toshiba drive.) Charles also sent results using FWB's HDT Benchtest on both the OEM 12GB Toshiba drive and the 32GB IBM drive. Although the 32GB Travelstar results don't look impressive compared to the 20GB Travelstar, compared to original equipment drives it's noticably faster. I wish I'd had an OEM drive to compare with the 32GB for my real-world tests in this PB G3/500. (As I did in my previous 20GB TravelStar review/guide in the PB Wallstreet.)
Finder/Applications Performance Tests:
Benchmarks usually show much higher transfer rates than real world tests which have file system overhead. I ran several tests on both the 20GB and 32GB Travelstar drives with exactly the same system folder, etc. on the drive. Since a stopwatch was used there is some human error in starting/stopping/repeatability, so times within 1/2 second or so are really a wash.
For the Finder duplicate folder/file tests I calculated the actual 'through-the-file-system' transfer rates (rounded to 1 decimal place). Pure benchmarks like ATTO Tools show much higher rates than Finder file copies, especially where a large number of files are involved which have a lot more overhead.
The boot time tests were from the startup tone until the Finder was ready to use (including mounting an expansion bay drive which was used to hold some test files). The startup RAM test was disabled.
Real World Performance Tests Test IBM 20GB Drive IBM 32GB Drive Boot Time 51.2 sec 51.2 sec Duplicate Folder
(457.2MB Q3 Pak file)
Launch Pshop 5.5 7.0 sec 6.5 sec Load 18MB PSfile 3.7 sec 3.0 sec Save 18MB PSfile 2.5 sec 2.5 sec Sherlock Search
(finding 4,500+ files)
6.8 sec 5.6 sec
I know it's strange to see the 4200 RPM drive perform so well compared to the 5411 RPM drive, but that's what I recorded. I regret I did not have a original PB drive to use for these tests, but for those considering which drive to buy between the 20GB and 32GB TravelStar models the above should be of some use.
Below are graphs of Macbench 5.0, ATTO Tools 2.3.2 and Intech Quickbench benchmark results of the IBM 20GB and 32GB Travelstar drives. (The ATTO Tools results also include one from Charles Eicher using the OEM Toshiba 12GB drive shipped in a PB Firewire G3/500. He also sent FWB Benchtest results between the 12GB Toshiba and the 32GB TravelStar.) Tests were run after a clean reboot to ensure as clean and consistent a system/ram/cache state as possible. For these tests only a system folder and a folder of benchmarks were installed on the drive.
MacBench 5.0: (Note: as many of you are aware, earlier this year MacBench was pulled from Ziff Davis' site and no longer supported. It's CPU test does not support Altivec and is highly influenced by cache size, etc. but I'm presenting disk scores using Macbench here regardless, since it is one gauge commonly used to measure disk performance and for comparisons to past disks tested with MacBench.) I ran the complete Macbench 5.0 disk test suite on both drives. Disk cache was set to the default OS 9.04 (auto) setting.
Note: As shown on the Benchmark page of my PB G3/500 Firewire review, the OEM 12GB Toshiba drive scored a 1383 on the Disk test and a 1007 on the Pub. Disk test, significantly lower than either of the new Travelstar drives.
ATTO Benchmark: The following are results of tests using ATTO's Tools Benchmark set to 8MB I/O size. First a summary of the results comparing the OEM 12GB Toshiba, 20GB TravelStar and 32GB TravelStar drives:
Since the graphs of rates across file size are also interesting, I've included them below for each of the 3 drives.
The 12GB Toshiba rates look low, but they're similar to what I saw in the Benchmarks page of my PB G3/500 Review.
Note: The reason the IBM drive graphs have an 'Extensions On' note is that I had tested with Extensions off and saw almost 10MB/sec better Peak Write (only) performance. (Granted the extension set included Palm extensions and Timbuktu.) What's odd was that Macbench disk and pub disk scores were lower with extensions off (by 31% on disk mix test and 19% on Pub disk test).
Intech QuickBench Tests: Intech's HD Speedtools drivers come with a benchmarking utility called QuickBench. The results of tests on the 20GB and 32GB Travelstar drives are shown below:
Results of these tests also show similar performance between the two drives.
FWB HDT 4.0 Benchtest: Although I have not used FWB's Benchtest for years now (I've seen results from it higher than even the drive mfr. claims), Charles sent HDT v4 Benchtest results on the OEM 12GB Toshiba vs IBM 32GB drives:
Note that although transfer rates were higher with the 32GB Travelstar, I/Os per second (Read/Write transactions) were much higher with the Toshiba drive.
General Drive Comments and Summary:
I have to admit I expected the 32GB 5411 rpm drive to perform better than it did against the 20GB 4200 rpm TravelStar model. However the 20GB TravelStar is an excellent performer and either drive is significantly faster than the original equipment drives.
If you want the largest hard drive possible in a notebook computer (to date), the IBM 32GN TravelStar is certainly the drive to buy. If you don't need appx. 30GB of storage however, consider the lower cost 20GB model which can save you a few hundred dollars at today's prices.
- In many tests the 32GB and 20GB TravelStar drives were literally identical (boot time, similar ATTO Tools performance, duplicate file tests, time to launch Photoshop, etc.)
- MacBench disk scores were a bit faster with the 32GB drive.
- Sherlock searches were faster with the 32GB drive.
- Compared to original equipment PB drives, the latest IBM TravelStars are significantly faster drives.
- Drive noise: The 32GB drive was quieter when the head was seeking/accessing data than the 20GB model, but there is a very slight (almost fan-like) noise when it's idle. I did not notice this with the 20GB (4200 rpm) model. (My sample may not be typical in this respect however.) Both drives are far quieter than any Toshiba drive I've used (both the review Pismo OEM Toshiba drive and an 18GB Toshiba drive I've used had a noticable whine.)
- Power Saving Modes: Although I have never seen the problem personally in daily use, one TravelStar 20GB owner reported a problem waking from sleep and noted he thought the drive 'clunked' a lot from the auto-head parking. I didn't find either of these a problem here but welcome other owners feedback.
The 32GB drive formats to just under 30GB (29.8GB). If you don't need the extra space, the 20GB model may be a better value since it sells for under $300, where the 32GB model sells for over $500 (based on current pricing as of Oct. 2000).
A special thanks to Transintl.com for sending a loaner for review.
Test System Details:
- PowerBook G3/500 Firewire
- 256MB RAM, VM Off
- Firmware v2.7
- OS 9.04 with all updates applied
(full extension set incl. Palm and Timbuktu, less OS 9 Multiple Users)
- Resolution Used: 1024x768, thousands colors
- PowerBook Wallstreet Drive Upgrade Guide
- IBM TravelStar 20GB Performance Tests (Compared to the OEM 4GB IBM drive in PB Wallstreet)
- Build Your Own Expansion Bay Hard Drive (guide to kit build)
- Portable firewire drive articles (including kit builds) tested with Powerbooks are on the Firewire articles page.
- See the Systems page Powerbook area, IDE articles pages for more Powerbook related reviews, guides and tips.
Index of PowerBook G3 Firewire HD Install Guide
= or =
Back to www.XLR8YOURMAC.com
Copyright © 2000 , All Rights Reserved.
No part of this sites content is to be reproduced in any form without written permission.
All brand or product names mentioned here are properties of their respective companies.
Site terms and conditions of use.