Abstract:
From virtualization to OpenGL GPU-multicast, we'll cover the full range of NVIDIA technologies available for implementing systems that contain multiple GPUs for real-time graphics applications. Techniques available for leveraging multi-GPU architectures range from fully implicit (e.g., GRID) to fully explicit (e.g., GPU affinity and multicast). We'll discuss the specific advantages of each approach, along with examples for how each method can be implemented. We'll also cover synchronization and streaming.
From virtualization to OpenGL GPU-multicast, we'll cover the full range of NVIDIA technologies available for implementing systems that contain multiple GPUs for real-time graphics applications. Techniques available for leveraging multi-GPU architectures range from fully implicit (e.g., GRID) to fully explicit (e.g., GPU affinity and multicast). We'll discuss the specific advantages of each approach, along with examples for how each method can be implemented. We'll also cover synchronization and streaming.
Back
Topics:
GPU Virtualization, Virtual Reality & Augmented Reality, Real-Time Graphics
Event:
GTC Silicon Valley