A Performance Comparison:
Apple Power Macintosh Computers vs. Intel Pentium Processor-Based Computers Running Windows

Executive Summary of an Independent Benchmark Study
by Competitive Assessment Services
July 1995

The latest PowerPC processor-based Power Macintosh computers from Apple Computer continue to significantly outperform comparable Intel Pentium processor-based personal computers running Microsoft Windows 3.1 or Windows for Workgroups 3.11. This finding is the result of a benchmark test program recently performed by Competitive Assessment Services, an independent test facility located in Huntington Beach, California.

The test program was conducted during July 1995. The program was designed to compare--quantitatively--the performance of a variety of Power Macintosh and Intel processor-based computers.

The study showed that Apple Power Macintosh computers consistently performed faster than similarly configured Pentium processor-based PCs in a wide assortment of tasks, including graphics, publishing, and scientific applications. Some key findings follow:

Power Macintosh 9500/132 Up to 123 Percent Faster

Power Macintosh 8500/120 Up to 98 Percent Faster

Power Macintosh 7500/100 Up to 71 Percent Faster

Power Macintosh 7200 Systems Up to 62 Percent Faster

The application-level testing was conducted on equivalently configured Power Macintosh and Pentium processor-based PCs running Windows. The tests measured the actual elapsed time required to perform various tasks. Unlike processor-only or low-level benchmarks, the test results reflect application-level performance running real applications on actual systems. The tasks involved a mix of integer, floating point, disk, and graphics activities.

About the Applications

The application-level benchmarks measured 58 different tasks performed in 10 different applications, including spreadsheet, word processing, database, document-layout, and business graphics applications. The specific applications were Microsoft Excel, Microsoft Word, ClarisWorks, Microsoft FoxPro, Adobe FreeHand, Fractal Design Painter, FrameMaker from Frame Technology, DeltaPoint DeltaGraph, Mathematica from Wolfram Research, and Ashlar Vellum.


The benchmark tests for each application comprised a sequence of user actions intended to accomplish standard measurable program functions or actions. The test sequence was divided into segments, each of which related to a specific program action, such as opening a file, scrolling, and recalculating a spreadsheet. Each segment was timed individually and a total time was compiled for the entire test case. All timed sequences measured only program and system activities; excluded from these sequences were times related to human reaction or required keyboard actions.

Each test series was performed at least three times from beginning to end. The average time of the three trials became the time for each step. Tests were designed so that timed segments were long enough to minimize the relative significance of timer-induced error.

Configurations and Results

Comparison Result Chart

(c) 1995 Apple Computer, inc. All rights reserved. Apple, the Apple logo, and Power Macintosh are trademarks of Apple Computer, Inc., registered in the U.S.A. and other countries. Adobe is a trademark of Adobe Systems Incorporated or its subsidiaries and may be registered in certain jurisdictions. ClarisWorks is a registered trademark of Claris Corporation. PowerPC is a trademark of International Business Machines Corporation, used under license therefrom. This document may not be reprinted without express written permission from Apple Computer, Inc., and Competitive Assessment Services.

- webmaster@www.apple.no -