SpeedTools QuickBench For MacOS X

QuickBench can measure eleven different data transfer sizes under four different scenarios: Sequential Reads, Sequential Writes, Random Reads and Random Writes. The results take into account four of the following influences: drive performance, driver performance, bus capability and your drive's onboard caching capability.

QuickBench provides two ways to view your benchmark results; in a dynamic graphing presentation or in a precise numeric presentation. In addition, the QuickBench utility provides a way to save your test results as a file which can be viewed and/or printed using Apple's SimpleText™ utility.

QuickBench is universal too, in that, it can measure more than just hard drives. QuickBench can be used to benchmark just about any mounted writable volume - like RAM disks! Additionally, it is compatible with any Macintosh model that supports "drag and drop"... even a Mac Plus!

The QuickBench utilities even offer Extended tests! Extended tests measure the speed of transfer sizes between 2 and 10 Megabytes. While this is far beyond what most drives will see with real applications, it nevertheless offers an interesting look at the maximum sustainable speed your drive and computer can achieve together within your current configuration.

Please note that QuickBench's Extended tests are very different from what normally happens when other applications read from and write to very large files. When most applications access very large files, they do so by breaking up the file transfer into smaller pieces. This dramatically reduces the amount of free RAM required to access very large files. For example, if an application program wanted to read in a 100 Megabyte file all at once, it would require 100 MB of free RAM. To save memory, the application may break up the transfer into a series of 100 transfers, one Megabyte for each. This method would require only one Megabyte of free RAM to read or write the entire 100 Megabyte file.

The QuickBench testing methodology is straight forward. The smaller test sizes of the Graphic and Numeric views use multiple iterations which are averaged together to produce an accurate and repeatable measurement for data transfers for each transfer size category. We do not believe anomalous spikes or dips in transfer rates, which are characteristic of smaller data transfers, tell us anything useful about given drive. Furthermore, we also do not subscribe to the notion of an overall "average" transfer rate, because this forces the software designer to make decisions about which measurements should and should not be included in the "average." Our experience is that sustained transfers are more useful to compare one drive variable against another.

On the other hand, the Extended tests are calculated based upon a single data transfer for each transfer size category. The Extended test is highly synthetic (meaning nobody ever really transfers data in chunks this large), but is a good way to measure the maximum rate which your device can sustain. Keep in mind that in almost every instance we know of, the device driver controlling the device you are testing will break up the Extended test transfers into multiple smaller transfers. The number of chunks that the driver breaks a given transfer into and the efficiency with which a driver does this will vary and can be a useful point of comparison between different device drivers which can be installed for a given storage device.
QuickBench Screen Shot
System Requirements

MacOS X (10.2 or later)
Download QuickBench User's Guide (pdf)