Parallel Nsight 2.0 and CUDA 4.0 for GPU Computing - SIGGRAPH 2011

Explore the powerful features of Parallel Nsight™ 2.0 and CUDA 4.0 in this SIGGRAPH 2011 conference talk. Dive into the CUDA parallel computing architecture and its ability to dramatically increase computing and graphics performance by leveraging GPU power. Learn about Parallel Nsight, the first development environment for CUDA and graphics application development integrated into Microsoft Visual Studio. Discover key features, setup processes, and practical applications through comprehensive demonstrations and explanations. Gain insights into Visual Studio integration, debugging tools, performance analysis, and optimization techniques for CUDA and graphics applications. Explore topics such as multi-system setup, synchronization, disassembly, memory copies, API level tracing, and SLI bridge functionality. Examine advanced features like single frame capture, render target inspection, pixel history, and targeted experiments. Understand how to leverage the latest hardware and CUDA support for optimal performance in your parallel computing projects.

Introduction
Overview
CUDA 40 Features
Parallel Nsight Overview
Features
Setup
MultiSystem Setup
How It Works
Setting Up Synchronization
Setting Up Local Directory
Visual Studio Integration
Visual Studio Disassembly
Debug Focus Window
New Analysis Activity
CUDA Launch Summary
Memory Copies
System Level
Filtering Threads
Do 3D perf markers
Do GPU draw calls
How many draw calls do you need
API Level Trace
Alien vs Triangle
SLI Bridge
Single Frame Capture
Saving Captures
Graphs
Geometry Units
CPU Side Statistics
Min Geometry
Composite
Render Target Inspection
Frames Page
Inspector
Navigation Bar
Shader Window
Pixel History
Pixel Experiment
Debug Pixel
Fragments
Pixel Picture
Exit Debugging
Profile Frame
Targeted Experiments
Syncing Data
State Buckets
Draw Calls
Dispatch Calls
Summary Page
Input Assembly
Frame Buffer
Wait Time
Unweighted
Primitives
Overview Graph
Sinking With The Target
Utilization Information
View All Graphics Resources
Graphics Profiler
Driver Independence
Latest Hardware Support
Latest CUDA Support