SLi

SLI, or Scalable Link Interface, is a multi-GPU technology developed by Nvidia that enables two or more graphics processing units to render a single image. The goal is to increase graphics performance and frame rates in supported games and applications by sharing rendering tasks across GPUs.

SLI achieves this through inter-GPU communication and driver scheduling. Rendering can be divided within a frame in one of several modes, most commonly Alternate Frame Rendering (AFR), where GPUs render alternating frames, and Split Frame Rendering (SFR), where the image is divided into spatial regions assigned to different GPUs. A physical SLI bridge (or newer interconnect) and a compatible motherboard are typically required, along with an SLI-capable graphics card and driver support. The effectiveness of SLI depends on software support and game-specific driver profiles, and performance gains are not guaranteed or linear.

Hardware and software requirements include a supported NVIDIA GPU lineup, an SLI-ready motherboard and chipset, an SLI bridge or equivalent interconnect, sufficient PCIe bandwidth, and up-to-date NVIDIA drivers. Not all games support SLI, and some may perform worse when enabled. In practice, multi-GPU scaling varies and can introduce latency and micro-stutter.

Historically, SLI originated from the Scan-Line Interleave technology developed by 3dfx in the late 1990s and was adopted by Nvidia in the mid-2000s. Over time, consumer adoption declined as single-GPU performance improved and software support waned, with Nvidia shifting focus toward other multi-GPU solutions such as NVLink for select products and professional workloads.