Reading Between The Threads: Shader Intrinsics

GameWorks, GameWorks Expert Developer, DX11, DX12, Vulkan, OpenGL, nvapi, DesignWorks

Mathias Schott, posted Jul 29 2016

Introduction

When writing compute shaders, it’s often necessary to communicate values between threads. This is typically done via shared memory. Kepler GPUs introduced “shuffle” intrinsics, which allow threads of a warp to directly read each other's registers avoiding memory access and synchronization. Shared memory is relatively fast but instructions that operate without using memory of any kind are significantly faster still.

Read more

Real-Time Graphics at Siggraph 2016

GameWorks

David Coombes, posted Jul 21 2016

There is a lot of cool stuff happening at Siggraph this year for game developers. The Advances in Real-Time Rendering course has become a must attend for graphics developers at the leading edge. Another key course is Open Problems in Real-Time Graphics which talks about the hard problems in computer graphics and how to solve them.

Read more