For testing today, we turned to IBM/Lenovo for their mid-range ThinkPad R52 laptop. The model we utilized for testing contained an Intel Pentium M 750, which is clocked at 1.86GHz, 90nm process, Dothan core, 2MB L2 Cache, 533MHz FSB, and supports EIST (Enhanced Intel SpeedStep Technology) with Execute Disable Bit. Below are the other components utilized in testing with the ThinkPad R52.
|Processor:||Intel Pentium M 750 (1.86GHz)|
|Motherboard:||Lenovo R52 18494WU (i915PM + ICH-6M)|
|Memory:||1 x 512MB DDR2|
|Graphics Card:||ATI RADEON X300 64MB (dedicated)|
|Hard Drives:||IBM 80GB 5400RPM|
|Optical Drives:||DVD±RW Drive|
|Add-On Devices:||Intel PRO/Wireless 2915ABG & Atheros AR5212|
|Operating System:||Ubuntu 5.10 (Breezy Badger) Preview|
|GCC (GNU Compiler):||GCC 4.0.2|
|Graphics Driver:||ATI RADEON 8.16.20|
In addition, the R52 utilizes the i915 Express mobile Chipset with ICH-6M, 15.1" TFT SXGA+ display (1400 x 1050), and AD1981B AC '97 audio. In testing today, the Pentium M CPU was ran at its stock speeds and voltages due to a lack of laptop adjustments and at this present time, we're only posting the results from the M-750 CPU but have intentions to bring additional results soon. The CPU cooler used with the Pentium M was left stock with no after-market cooling. Although the laptop shipped with Microsoft Windows XP Professional, part of the reason we chose to use the laptop was its superior Linux support. Sticking to the niche of our favorite Linux distributions for testing and benchmarking purposes, we went with Ubuntu Linux 5.10 Preview due to its bleeding edge status when it comes to the kernel, GNOME, GCC, and various other packages.
For benchmarking today, we have skipped out on our usual gaming OpenGL benchmarks due to the slow ATI X300 it utilized with poor drivers, and instead went for more CPU/memory centric programs. What we settled for using in this Pentium M 750 Linux preview is LAME Compilation, LAME Encoding, Gzip Compression, BlueSailSoftware Opstone benchmarks, and FreeBench. For additional testing, although not directly CPU related, we also ran RAMspeed. If you are unfamiliar with our traditional Phoronix benchmarking practices, we would recommend you read this page for many of our standardized procedures, or read one of our other many recent benchmarking articles. All benchmarks in the testing were ran three times, with the average of the three runs being displayed except where otherwise stated on the results page with the values from all trial runs.