2.2 Acceleration Features
•
Support for DirectX® 12 (Feature Level 12_0) features, including the full-speed
32-bit floating point per component operation:
•
Shader Model 5.0 geometry and pixel support in a unified shader
architecture:
∘
Vertex, pixel, geometry, compute, domain, and hull shaders.
∘
32- and 64-bit floating-point processing per component.
∘
New advanced shader instructions, including flexible flow control
with CPU-level flexibility on branching.
∘
A nearly unlimited shader-instruction store, using an advanced
caching system.
∘
An advanced shader design, with an ultra-threading sequencer for
high-efficiency operations.
∘
Graphics Core Next supporting native scalar instructions.
∘
Advanced, high-performance branching support, including static
and dynamic branching.
∘
High dynamic-range rendering with floating-point blending, texture
filtering, and anti-aliasing support.
∘
16- and 32-bit floating-point components for high dynamic-range
computations.
∘
Full anti-aliasing on renderable surfaces up to and including 128-bit
floating-point formats.
∘
A new read/write caching system, replacing texture cache with a
unified read-write two-level cache.
•
Support for OpenGL 4.5.
•
Support for OpenCL™ 2.0.
•
Support for Mantle
•
Support for AMD LiquidVR™
•
Anti-aliasing filtering:
•
2×/4×/8× MSAA (multi-sample anti-aliasing) modes are supported.
•
A multi-sample algorithm with gamma correction, programmable sample
patterns, and centroid sampling.
•
Custom filter anti-aliasing with up to 12-samples per pixel.
•
An adaptive anti-aliasing mode.
•
Lossless color compression (up to 16:1).
Functional Overview
5
©
2017
Advanced Micro Devices, Inc.
AMD Confidential - Do not duplicate.
"Vega 10" Databook
56006_1.00