Classic MacOS Emulation Performance
Objective: test real-world performance of MacOS 9.1 in native and emulated mode under MacOS X and PowerPC Linux/MOL.
Hardware: Apple iBook FireWire (Summer 2000) model, 366 MHz PowerPC G3 CPU, 256 K L2 Cache, 320 MB RAM, 10 GB HD internal IBM 4200 rpm HD, partitioned - 6 GB for HFS+ MacOS and the rest for Linux. Although iBook with relatively slow hard drive is not the best system for such tests, I just did not had another hardware with MacOS 9.1, MacOS X and Linux installed.
MacOS Classic Software: real-world MacOS 9.1 (e.g. with all my extensions, including ATM), extensions on, virtual memory on (RAM size + 1 MB), 1500 KB disk cache, AppleTalk on, file sharing off.
MacOS X Software: MacOS X v10.0.3, installed on the same HFS+ formatted partition as MacOS 9.1, AppleTalk on, File Sharing off.
Linux Software: SuSE Linux 7.1, kernel 2.4.2, GLIBC 2.2, single ReiserFS partition, 96 MB swap partition, MOL (Mac on Linux) emulator 0.95.8-1 running in X (slow video) mode, 160 MB of RAM allocated for MacOS 9.1 running under MOL, running services - Apache, Postfix, CUPS, NFS, Netatalk, XFree86.
Benchmark Software: MacBench 5.0. Although believed to be outdated, it still allows to measure relative CPU/HD/2D Graphic performance of Classic MacOS.
Notes: MacOS under MOL loaded with the same extensions, virtual memory set to 160 MB + 1 MB because half of RAM have been used by Linux and MOL. In custom tests block size have been equal to 4096K. Of course, all disk test performed on the same target MacOS HFS+ partition. In all test larger bars are better.
Special Notes: These tests do not measure native MacOS X performance! They are applicable to MacOS 9.1 running in native mode and under different emulation conditions only.
The results are nothing but big surprise. Under emulation CPU and FPU scores have been little lower as expected. However, MacOS running under MOL has considerably higher disk scores (up to 8 times in some tests)! I repeated test several times, but performance scores always have been very close. I took another disk benchmarking utility called QuickBench, a part of Intech Hard Disk Speed Tools (which is supplied with retail version of many third-party hard drives). Unfortunately, QuickBench does not work under MacOS X.
Take a look at disk data transfer rate throughput. Linux layer offers almost tripled performance edge over plain MacOS 9.1 running in native mode.
The question arise is it possible to boost Classic MacOS performance and make it more usable with either MacOS X either Linux. Unfortunately, in almost all cases not. For example, MOL 0.9.58-1 does not support video acceleration. Screen refresh of Classic under MacOS X seems to be slower, too. MacOS performance under MOL significantly drops if Linux begin to do something disk and/or CPU intensive (e.g. backup with compression).
These benchmark tests actually allow to reveal weak points of Classic MacOS and judge possible performance gain with better optimized software.