Nvidia NVIDIA Quadro Manuel D’Utilisation

Page de 9
 
 
SLI for NVIDIA Quadro: Installation Guide 
 
This document provides the information necessary to successfully install and configure your SLI-
capable graphics boards and system.  You may also review the full Installation Guide for 
additional information.  Topics covered in this document include: 
 
 
• 
NVIDIA SLI technology 
 
• 
Motherboard and power supply requirements 
 
• 
SLI hardware setup and configuration 
 
• 
Display driver installation and enabling SLI 
 
• 
SLI questions and answers 
 
NVIDIA SLI Technology 
NVIDIA SLI (Scalable Link Interface) technology allows users to install and benefit from two PCI 
Express (PCIe) graphics boards in a single PC.  Just as dual processors or dual core CPUs 
improve compute performance, SLI allows dual GPUs (Graphics Processing Units) to significantly 
accelerate graphics performance.  Professional users benefit by using SLI technology on dual 
NVIDIA Quadro GPUs to span an OpenGL window across multiple displays, or to run a single 
application per GPU. Additionally, the SLI connector offers dual GPU performance on a single 
display by leveraging the capabilities of two NVIDIA Quadro GPUs. 
 
SLI for NVIDIA Quadro – Definitions 
SLI Frame Rendering: Combines two PCI Express graphics boards with an SLI connector to 
transparently scale application performance on a single display by presenting them as a single 
graphics card to the operating system. 
 
Benefits: 
- Visual simulation, broadcast, and video applications that are fill rate-limited will benefit from 
using the split frame rendering (SFR) mode.  
- SPECviewperf® and other applications that are geometry-limited will benefit from using the 
alternate frame rendering (AFR) mode. 
 
Requirements:  
- Two identical NVIDIA Quadro FX boards (4400, 3400, 1400)  
- SLI connector 
 
SLI Multi View: Combines the power of two NVIDIA Quadro PCI Express graphics boards to span 
a single hardware-accelerated OpenGL application window across multiple displays, run a single 
application per GPU with multiple display outputs, or enable other flexible usage of two PCI 
Express graphics boards.  
 
Benefits: 
- View professional applications over multiple displays to increase visual real estate.  
- Run a single application per GPU to offload geometry processing to a second GPU.