NVIDIA Linux SLI Primer
Now that the 1.0-8XXX drivers are due out for an official public launch shortly, and the 1.0-8168 BETA drivers have been leaked by ASUS, GNU/Linux users can finally begin to experience the benefits of Scalable Link Interface (SLI) Technology in addition to many other improvements that Microsoft Windows ForceWare users have been able to utilize in their high-end systems for quite some time. For those Linux users who aren't familiar with this SLI Technology, rather than running a single graphics card, you're able to run two identical graphics cards in tandem on supported PCI Express systems. Similar to Scan Line Interleaving found on 3DFX Voodoo2 3D accelerators, Scalable Link Interface is able to utilize two PCI Express x16 graphics cards simultaneously but rather than having each card render one line, NVIDIA uses a dynamic load balancing logic to split the rendering portions equally between the two GeForce 6/7 units. For Linux users today we're going to discuss the possible options for establishing a NVIDIA SLI system when it comes to the motherboard, graphics card, and power supply. We also have a few tidbits of information (with SLI screenshots) when it comes to the SLI drivers and software at this point in time.
Before we begin, the article we have here today is simply a guide for gaming enthusiasts looking to construct their own NVIDIA SLI (Scalable Link Interface) system running Linux, and we aren't including any graphics benchmarks with this piece. However, we have numerous articles lined-up for posting upon the official 1.0-8XXX launch, which is expected in the near future.