Archive Home

Diamond Viper ATI Radeon HD 3870 512MB Video Card E-mail
Reviews - Featured Reviews: Video Cards
Written by Ronald Tibbetts   
Tuesday, 11 March 2008
Table of Contents: Page Index
Diamond Viper ATI Radeon HD 3870 512MB Video Card
HD 3870 Package and Contents
HD 3870 Technology
Features and Specifications
Diamond HD 3870: Closer Look
HD 3870: Detailed Features
Video Card Testing Methodology
Real World Benchmarks
Synthetic Benchmarks
Final Thoughts and Conclusion

Video Card Testing Methodology

Testing Methodology:

For testing we used an Intel E6600 as our base processor as it's a CPU that reflects a vast majority of Core 2 Duo systems at the present. With its wide range of multipliers (6x-9x) the E6600 can easily can achieve 3.0 Ghz at a FSB of 333. The E6600 has a stock FSB of 1066; though compatible with a 1066FSB the ASUS P5K3 Deluxe motherboard we used for testing has been designed for a FSB of 1333. To remedy this we raised the FSB to 333 leaving the multiplier and voltage at default. This is essentially a free 25% overclock anyone with this set-up can benefit from, allowing everything in the system to run at stock values. Placing our mainstream E6600 CPU on par with the more expensive E6850's performance.

For our operating system (OS) we chose a fresh install of Windows XP with Service Pack 3 RC1 (SP3). Microsoft's Service Pack 3 rolls out some new features; however, it is essentially a cumulative update for all of Microsoft's hotfixes and updates. Although we had the option to test on a newer OS, we wanted the results to reflect better with real-world users, and according to a recent survey more than 90% of users still run Windows XP. Since most gamers and enthusiasts are still using Windows XP, it was decided that DirectX 9 would be used for all tests.

Every test was conducted at the following resolutions: 1920x1200 (24" widescreen LCD) 1600x1200 (20.1/21" Standard LCD's), 1280x1024 (19" Standard LCD), and 1024x768 (17" Standard LCD). Each test was run after a system restart, and the first benchmark for each test omitted. This process proved extremely relevant to testing, especially in games such as World in Conflict and Crysis, as the first run served to cache maps; with subsequent tests performing better than previous ones. Each test was run in succession three times to gain an average reproducible result.

Each benchmarks used was selected to represent a different gaming engine technology and graphic rendering processes, and we feel that this battery of tests will best serve to provide the diverse range needed for gauging performance.


Using GPU-Z (a utility available for free from our affiliate website techPowerUp!) we can verify the internal specifications for the Diamond HD 3870.

Benchmark Applications

  • 3DMark03
  • 3DMark05
  • 3DMark06
  • Lightsmark 07
  • Cinibench Release 10
  • Call of Duty 4 - Modern Warfare
  • World in Conflict v1.0.0.5 Performance Test (Very High Setting)
  • Crysis v1.1.1.5879 (High Settings, No Anti Aliasing)
  • Half-Life 2 Episode Two with Steam build date Feb. 07, 2008
  • Supreme Commander - Forged Alliance

Test Setup:

  • Processor: Intel E6600 Core 2 Duo 2.44GHz ~1.3v (9x333MHz FSB) 3.0Ghz
  • Cooling: Zalman CNPS8700 LED
  • Memory: 2 GB (1,024 MB x2 modules) Patriot PC3-15000 DDR3 @ 1333Mhz
  • Video: Diamond Multimedia VIPER HD 3870 512MB @ Stock w/ ATI Catalyst v7.12
  • Audio: Sound Blaster X-Fi
  • Hard Disk: 2xWD RAPTOR 74GB 16MB Cache Raid 0 10,000rpm, 1x WDC WD4000AAJS-00YFA0 (372 GB, IDE)
  • Optical Drive: SONY DVD RW DRU-830A SCSI Cd-Rom Device, HL-DT-ST DVD-ROM GDRH10N (16x/48x DVD-ROM)
  • PSU: Ultra X3 1000 watt modular cables
  • Operating System: Windows XP Professional Service Pack 3, v.3264 with IE7 (OS Version 5.1.2600 WinXP Retail)

Test Tools


Comments have been disabled by the administrator.

Search Benchmark Reviews

Like Benchmark Reviews on FacebookFollow Benchmark Reviews on Twitter