lors; for mainstream Windows users who use general-business applications.
-- 32-bit color:
1024- by 768-pixel resolution with 16.7 million (32-bit) colors; for users who are willing to sacrifice a small amount of performance for true color.
-- High resolution:
1600- by 1200-pixel resolution at 256 colors; for desk-top publishing, image processing, and phot
o-realistic rendering applications that require razor-sharp resolutions.
-- Low cost:
The same parameters as best overall but with a $400 price ceiling; for users on a budget who don't require higher resolutions and cutting-edge screen-refresh rates.
To be included for testing, we required that the adapters supported a minimum of 1024 by 768 pixels with 256 colors in noninterlaced mode. They had to have a minimum of 1 MB of video memory, but if an adapter could support a larger amount of RAM, we asked the vendors to supply the greater amount.
We tested the boards in a Gateway 2000 P5-120, a 120-MHz Pentium system with 16 MB of RAM, a 1-GB IDE hard drive, and a 21-inch NEC XP21 monitor. We installed a plain VGA version of Windows 95 and then created a disk image so that the original configuration could be restored each time we switched adapters to allow for a more real-world configuration procedure.
Performance is the most important thing to consider when buying a gra
phics accelerator, so that was our primary criterion for selecting the winners. After we chose the top performers, we ranked the winners and runners-up by considering the cost (for the low-cost category), support options, usability, and unique features that the boards offered.
PERFORMANCE
To get performance scores, we tested the adapters under real-world conditions. To do this, our graphics
tests use images produced
from widely used applications, such as CorelDraw, Corel Presents, Excel, PowerPoint, and Word for Windows. We also required each adapter to draw a variety of Windows images, ranging from straight text, to 2-D and 3-D bar charts, to complex full-color drawings.
To guard against cheating, we designed our tests to be resilient to benchmark-optimized drivers. And, to increase the tests' accuracy, we used microsecond timing. This lets us precisely measure a single screen paint, and it avoids the necessity of drawing the same screen repeatedly (which is
unrealistic and also easy to optimize in the driver).
Our Windows tests drew each of the 15 Windows application screens into both system memory and video memory using four different color modes for more than 120 tests. We also measured the time it took to refresh the screen from an image cached in memory at screen depths of 1, 2, 4, 8, 16, and 32 bits per pixel. Well-written applications cache display images whenever possible to improve response times.
FEATURES
Participating vendors supply us with responses to a lengthy survey that covers a wide range of graphics-adapter features. Some important features that we weight are the maximum resolutions and colors supported, the highest vertical-refresh rates, the number of adapters supported, and the different software-utility modules supplied with the board. We verify these responses and assign a rating based on the number of hits the board makes in the weighted categories.
USABILITY
In the ease-of-use score
s, an excellent ranking was reserved for adapters with exceptionally clear and complete documentation and installation software. Adapters received a good rating if an average user could install them without referring to the manual. Boards that were rated fair required our testers to consult the documentation, and poor ease-of-use scores indicate that we had to reset some setting or jumper or consult the vendor's technical-support personnel.
Although ease of installation was a factor in our usability scores, this judgment was tempered somewhat by the fact that once you get even the most troublesome board up and running, you're likely to be concerned only about performance and compatibility for the rest of the board's life. The cards in each category are weighted 75 percent for appropriate performance results, 15 percent for features, and 10 percent for usability.
Contributors
Jim Kane, Project Manager/NSTL,
h
as been testing network and PC hardware at NSTL for the past six years.
John McDonough, Technical Editor/NSTL,
has been writing for high-tech publications for several years.
Bruce Levy, Ph.D., Manager, R&D/NSTL
Stanford Diehl, Director of Product Reviews/BYTE
Helen E. Holzbaur, Manager, International Licensees and Methodologies/NSTL
Susan Colwell, Technical Editor/BYTE
screen_link (98 Kbytes)
One of the CorelDraw images NSTL uses to gauge response times for screen draws and redraws as part of the InterMark benchmarks.
How We Tested
