After careful consideration I have decided to transfer all hardware review activities to a new domain. I purchased Hardwareasylum.com in 2012 and have been working hard to build a new and improved Ninjalane on that domain. If you are reading this you have reached one of the archived articles, news, projects and/or reviews that were left behind during the site migration.
Please update your bookmarks and be sure to visit the new and improved Ninjalane at Hardwareasylum.com
Using multiple video processors isn't a new idea; 3dfx pioneered the process with the Voodoo2. 3dfx SLI, or rather "Scan-Line Interleave", allowed 2 or more Voodoo2 cards to be linked together which almost doubled overall video performance. AGP saw the first dual processor video cards both from 3dfx and ATI, however bandwidth and processing power were limited by the single AGP video connection. It wasn't until the advent of PCI Express where we saw the re-emergence multiple graphics cards and a new leap in video performance.
PCI Express is a high speed serial replacement for older parallel technologies like PCI and PCI-X. The new architecture also replaced AGP as the primary path for video processing due to the higher bandwidth and modular design. The simplest way to think of PCI Express is to describe the notion of "Lanes". Each lane is a single full duplex signal path than can carry between 250 and 500 MB/s of data. These lanes can then be grouped together in factors of 2 (1x, 2x, 4x, 8x, 16x) up to a maximum of 32x.
This article will be centered on the "factors of 2" limitation and how it translates into basic motherboard design and layout. The results we gather will be used to create the first ever Ninjalane Multi-GPU Index.