GSV3100 is a graphics processor IP conformant to OpenGL ES2.0/1.1 and OpenVG1.1. High performance, small footprint, and low power consumption – The GSHARK-TAKUMI architecture achieves all these at a higher level and is perfect for use in mobile/embedded system devices.
GSV3100 combines 3D rendering functions powered by the programmable shader and an OpenVG hardware accelerator intended for embedded system use.
Shader architecture type 3D GPU Integrating the OpenVG 1.1 hardware processing pipeline
Overview
Key Features
- Pixel performance
- 400M pix/sec
- Texture size
- Up to 8192 x 8192 pixels
- 3D rendering
- Integrated high-performance programmable shader:
- Unified shader and multiple thread processing to realize the fast and efficient parallel computes
- High-precision computing on the parallel compute unit in 32-bit floating point
- Transform, lighting, setup, rasterize, texturing, texture filtering, blending, dithering, full-scene antialiasing, vertex array engine
- 2D rendering
- Vector graphics engine
- Path processing, transform, rasterize, lines, curves (Quadratic/Cubic Bezier curves), elliptical arcs,
- paint (color/gradient/pattern), scissoring, alpha masking, alpha blending, antialiasing, color space conversion, image drawing (scaling/rotation/transformation), image filtering
- 2D Sprite engine
- Scaling, rotation, transformation, BitBLT, line/dot/rectangle drawing, blending, transparent color
- Other functions
- Copy engine (fast rectangle-fill rendering, format conversion)
- Standard conformance
- OpenGL ES 2.0/1.1 and OpenVG 1.1
Benefits
- Proprietary architecture – MATTS developed by TAKUMI
- GSV3100 adopts the MATTS architecture integrating the programmable shader that has successfully optimized a dynamic load balance of vertex and pixel (fragment) processing, combined with high performance and low power consumption at a time.
- Graphics IP core – Advanced and hybrid
- By integrating a native OpenVG accelerator, this high-performance, real-hybrid graphics IP core supports both 3D renderings on OpenGL ES 2.0 and vector graphics renderings on OpenVG 1.1.
- Footprint & power consumption – Small and low
- Hardware acceleration IP – High-performance/quality graphics rendering with minimized CPU load
Deliverables
- Synthesizable RTL code
- OpenGL ES2.0/1.1?OpenVG1.1 Library
- Associated documentation
Technical Specifications
Availability
Work in progress
Related IPs
- 3D OpenGL ES 1.1 GPU (Graphics Processing Unit)
- 2D (vector graphics) GPU IP Further advanced architecture for minimized CPU load and increased pixel performance in vector processing
- Third-Generation Valhall-based Graphics Processing Unit (GPU) for the sub-premium market
- Second-Generation Valhall-based Graphics Processing Unit (GPU) for the Mainstream Market
- 3-D Audio Processing Core
- Instruction Set Compatible with the 8051 8-bit Microcontroller Architecture