1. Computers
  2. Display Drivers
  3. Graphics Cards
  4. Memory
  5. Motherboards
  6. Processors
  7. Software
  8. Storage
  9. Operating Systems


Facebook RSS Twitter Twitter Google Plus


Phoronix Test Suite

OpenBenchmarking.org

Intel Centrino DDR2 Investigation

Michael Larabel

Published on 19 December 2005
Written by Michael Larabel
Page 1 of 6 - Comment On This Article

After Corsair's article earlier this year with their internal memory benchmark results on the performance impact of 1GB system memory versus 2GB, there have been countless articles from other publications, including Phoronix with our Linux 2GB testing, as to the true performance benefits from the expansion. Following suit over recent months has also been numerous manufacturers coming to the table with 2 x 1024MB kits for enthusiasts, gamers, and overclockers alike to replace the standard 2 x 512MB configuration. While 1024MB of system memory remains suitable for most applications, 2048MB is quickly becoming the de facto standard for those seeking the maximum real world benefits from their computer system. Nevertheless, this is a never-ending race as we still faintly recall the days of when a few Megabytes of system memory was considered overkill. However, in this mad memory upgrading rush how do the mobile systems fair with these memory upgrades? With one of our loaded Centrino systems in hand, we will see how these large memory capacities carry over to the mobile front. As DDR2 is quickly becoming the standard for system memory with all of Intel's latest i9XX series Chipsets taking advantage of the additional bandwidth, and AMD's next-generation CPUs are going to offer a DDR2-667 memory controller, we felt it wise to deliver our memory results using DDR2-533. Of course, the memory interface for notebooks is quite different with its 200-pin DDR2 SO-DIMM design compared to the 240-pin desktop alternative. The reason for this DDR2 memory take over is due to DDR reaching the end of the dock as following the standards set fourth by JEDEC, DDR can no longer scale any further, when it comes to non-overclocked frequencies. Although the mainstream i915GM and i915GMS Express Chipsets support the use of Double Data Rate Two, we turned to the Intel's only performance oriented Chipset that supports the use of DDR2 memory and that is their Mobile i915PM Express. Implemented with the i915PM is support for DDR2-400/533MHz system memory in either single or dual channel modes. Also powered by the Chipset is PCI Express x16 graphics, Enhanced Intel SpeedStep Technology, Intel High Definition Audio, and the 533/400MHz Pentium M series. For investigative purposes we will be benchmarking the Centrino notebook with 512MB, 1GB, and finally 2GB of DDR2 system memory, which is the maximum capacity supported by the i915PM.

For the 512MB implementation, we will be using a single stick of Elpida's PC2-4200S-444. The Japanese 512MB memory stick uses Elpida E5116AB-5C-E chips in a double-sided 1 x 4 configuration with a batch number of 05220WPPK. The Elpida sticker is labeled EBE52UD6ABSA-5C-E and 0529AXG0601. For the 1024MB and 2048MB memory testing, we turned to the talented engineers over at OCZ Technology. To carry out the testing we chose a pair of OCZ's 1GB OCZ25331024VSO notebook memory modules. These modules are designed for CAS 4 operation, like a majority of the DDR2-533 notebook memory modules, and offers comparable timings of 4-4-4 with IBM's Elpida module. In addition to the DDR2-533MHz CAS 4-4-48 operation, the modules run at 1.8V and come with OCZ's lifetime warranty. Examining the OCZ modules, used onboard are Elpida's E5108AE-5C-E ICs. These ICs are the same parts as what we had found several months ago when reviewing Transcend's TS64MLQ64V5J 2 desktop DDR2 memory modules. To offer the 1GB module capacity, on each side of the PCB are the Elpida chips in a 2 x 4 array. The sticker on each module simply displays the speed (PC2-4200), type (1024MB SODIMM), series (Value), timings (4-4-4-8), and part number (OCZ25331024VSO). These two OCZ modules were not part of a kit and the ICs on each of the modules came from a different batch, 0529A9B0C and 0531A9B2H. As other 2 x 512MB parts had not arrived on time for this article, the 2 x 1GB memory configuration is the only area that will be able to benefit from dual channel technology.

In need of the Intel i915PM Express Chipset to maximize the DDR2 notebook memory performance, we turned to our faithful IBM/Lenovo ThinkPad R52 that we had coincidently used in our mobile cache comparison article that was published yesterday. Inside the notebook is an Intel Pentium M 750 processor, which features a 90nm manufacturing process, 2MB of L2 cache, and is clocked at 1.86GHz with a 533MHz FSB with Enhanced Intel SpeedStep Technology and Execute Bit Disable support. Powering the graphics needs is a respectable ATI X300 with dedicated video memory. The ThinkPad R52 has two 200-pin DDR2 SODIMM memory slots with one of them being located on the bottom side of the notebook while the additional slot is located underneath the keyboard.

Hardware Components
Processor: Intel Pentium M 750 (1.86GHz)
Motherboard: IBM R52 18494WU (i915PM + ICH-6M)
Graphics Card: ATI RADEON X300 64MB (dedicated)
Hard Drives: Toshiba Slim MK1032GAX
Optical Drives: DVD-RW Drive
Add-On Devices: Intel PRO/Wireless 2915ABG
Software Components
Operating System: FedoraCore4
Linux Kernel: 2.6.14-1.1653_FC4
GCC (GNU Compiler): 4.0.0
Graphics Driver: ATI v8.20.8
Xorg: 6.8.2

Loaded on the notebook was BIOS version 1.22 (76ET62WW) and the embedded controller version was v1.04. The benchmarks running on this system was our traditional set that consists of Enemy Territory, Doom 3, Quake 4, LAME Compilation, LAME Encoding, FreeBench, and finally RAMspeed. As you can see from our arsenal of traditional Linux memory benchmarks there is a great mixture of various games as well as applications to display the overall real-world system performance. During testing, all of the memory modules ran at their designed DDR2-533 speeds at 4-4-4-12 timings. On the following pages are our results from this investigation.

Latest Linux Hardware Reviews
  1. CompuLab Intense-PC2: An Excellent, Fanless, Mini PC Powered By Intel's i7 Haswell
  2. From The Atom 330 To Haswell ULT: Intel Linux Performance Benchmarks
  3. AMD Radeon R9 285 Tonga Performance On Linux
  4. Apotop Wi-Copy
Latest Linux Articles
  1. AMD Moves Forward With Unified Linux Driver Strategy, New Kernel Driver
  2. MSI: Update Your BIOS From The Linux Desktop
  3. NVIDIA vs. AMD 2D Linux Drivers: Catalyst Is Getting Quite Good At 2D
  4. 15-Way GPU Comparison With Mesa 10.3 + Linux 3.17
Latest Linux News
  1. NVIDIA's NVPTX Support For GCC Is Close To Being Merged
  2. KDE's KWin On Wayland Begins Using Libinput
  3. Khronos Releases OpenVX 1.0 Specification
  4. Linux Kernel Working Towards GNU11/C11 Compatibility
  5. Ubuntu 15.04 Is Codenamed After A Monkey: Vivid Vervet
  6. Following GCC, Clang Looks To Default To C11
  7. Users/Developers Threatening Fork Of Debian GNU/Linux
  8. Linux 3.18-rc1 Released One Week Early With Many Changes
  9. The VC4 Gallium3D Driver Is Still Moving Along For The Raspberry Pi
  10. Direct3D 9 Support Might Land Within Mainline Mesa 3D Drivers
Latest Forum Discussions
  1. HOPE: The Ease Of Python With The Speed Of C++
  2. Bye bye BSD, Hello Linux: A Sys Admin's Story
  3. Users/Developers Threatening Fork Of Debian GNU/Linux
  4. NVIDIA Presents Its Driver Plans To Support Mir/Wayland & KMS On Linux
  5. AMD Is Restructuring Again, Losing 7% Of Employees
  6. Open-Source AMD Fusion E-350 Support Takes A Dive
  7. Upgrade to Kaveri, very slow VDPAU performance
  8. ChromeOS Drops Support For EXT2/EXT3/EXT4 File-Systems