Preparing For And Running SYSmark 2004 SE : How We Test Laptop and Notebook Computers for Home and Office

Preparing For And Running SYSmark 2004 SE

By Barry Gerber, published on July 19, 2005
Source: Tom's Guide US | Keywords: how, we, test, laptop, and, notebook, computers, for, home, and, office | Themes: Business Notebooks

Contents

1. Testing For The Rest Of Us
2. Different Kinds Of Mobile Computer Users
3. Battery And Performance Benchmarks
4. Preparing For And Running MobileMark 2005
5. Preparing For And Running MobileMark 2005, Continued
6. SYSmark 2004 SE (Second Edition)
7. SYSmark 2004 SE (Second Edition), Continued
8. Preparing For And Running SYSmark 2004 SE

8. Preparing For And Running SYSmark 2004 SE

Except for the system, Windows and other settings discussed in the table under MobileMark 2005, we make the same preparations for SYSmark 2004 that we make for MobileMark 2005. We configure BIOS and Windows parameters to optimize performance for a laptop or notebook running on AC. Here are the settings we use.

BIOS
USB and all device ports	All enabled
Wireless and wired network adapters	Disabled (if cannot be disabled in BIOS, do it in networking software: Control Panel > Network Connections, select and right-click network connection and select Disable)
Windows XP
Control Panel > Power Options > Power Schemes > Power Schemes dropdown	Home/Office Desk
Control Panel > Power Options > Power Schemes > Turn Off Monitor	Never for AC
Control Panel > Power Options > Power Schemes > Turn Off Hard Disks	After 20 minutes for AC
Control Panel > Power Options > Power Schemes > System Standby	Never for AC
Control Panel > Power Options > Power Schemes > System Hibernates	Never (this option is not shown if Control Panel > Power Options > Hibernate > Enable Hibernation is unchecked, see below)
Control Panel > Power Options > Hibernate > Enable Hibernation	Disabled (Unchecked)
Audio	Enabled
Pointing device settings	Default
Control Panel > Display > Settings > Screen Resolution	1024 x 768 minimum
Control Panel > Display > Settings > Color Quality	65536 colors minimum
Control Panel > Display > Screen Saver > Screen Saver dropdown	None
Control Panel > Display > Appearance > Color Scheme	Windows and Buttons: Windows Classic Style; Color Scheme: Windows Standard; Font Size: Normal
Active Desktop (Control Panel > Display > Desktop > Customize Desktop > Web	Uncheck all items
Control Panel > System > Advanced > Performance > Settings > Visual Effects	Adjust for best performance
Control Panel > System > Advanced > Performance > Settings > Advanced > Processor Scheduling	Set for Best Performance of Programs
Control Panel > System > Advanced > Performance > Settings > Advanced > Memory Usage	Set for Best Performance of Programs
Control Panel > System > Performance > Performance Options > Advanced > Virtual Memory > Change	Set to double the amount of RAM in the system
Control Panel > System > System Restore > Turn off System Restore On All Drives	Check to Disable System Restore
Control Panel > System > Remote > Allow Remote Assistance Invitations To Be Sent From This computer	Uncheck
Control Panel > System > Remote > Allow Users To Connect Remotely To This Computer	Uncheck
Control Panel > System > Automatic Updates > Turn Off Automatic Updates	Select

Once we have completed all of our preliminary settings, we are ready to execute a SYSmark 2004 SE trial run. This ensures that all of the scripts complete without a problem, and optimizes Windows XP settings; for example, the disk drive that SYSmark 2004 SE runs on is defragmented. Even if the trial run fails for some reason, the disk is defragmented.

With the trial run under our belts, we run our benchmark tests. As with MobileMark 2005, the SYSmark 2004 SE interface makes it easy to run tests and access the results. SYSmark 2004 SE scripts include a good deal of multitasking, with the focus on applications changing periodically from one window to another. The scripts are more likely to fail on laptops and notebooks due to a loss of window focus than MobileMark 2005 scripts. Thus, we find ourselves rerunning SYSmark tests more often than MobileMark tests; we often cannot run both benchmark scenarios together, requiring that we compute an unofficial overall score for the two scenarios. Update (10/05/05): We now understand both MobileMark and SYSmark better and most of our test runs are successful.