Производительность шейдера обработки пикселов (GFLOPS)
16 пикселя в секунду
Частота шейдера (МГц)
1000
Производительность видеопроцессора
123 миллиона треугольников в секунду,
800 миллионов пикселей в секунду.
Вершинный шейдер
2
Производительность вершинного шейдера
500 миллионов вершин в секунду
Кеш памяти 2 уровня (Level 2 cache)
Технологический процесс
0,045 микрон
DirectX
11 (9_3) SM3.0
Khronos API Support
OpenGL ES 1.1/2.0
OpenGL ES 3.0 — опционально
OpenGL 3.0/2.1
OpenVG 1.1
OpenCL 1.1 — опционально
WebGL
Renderscript Compute (включая Filterscript)
Особенности
GPGPU Options 1.1/1.2
Cache Coherent (Последовательный кеш)
Heterogeneous System Architecture (HSA) ready to accelerate workload in parallel with multi-core CPUs, DSPs, and FPGAs
Cores scalable from tens of GFLOPS to over 1 TFLOP in various multi-core GPU versions
Dynamic scheduler and command dispatch unit for optimized shader load balancing and resource allocation of millions of parallel
Compute threads (thread management with no software overhead)
Hardware context switching between threads hides memory latency so core is always working on active data
Ultra-threaded, unified shaders maximize Compute processing throughput and minimizes latency
Plug-N-Play integration – Stream Interface for FPGA fabric, standard ACE-Lite (cache coherent), AXI, and AHB interconnects
Fast, immediate hidden surface removal (HSR) that eliminates render processing time by an average of 30%
Extreme low power – Built-in battery-saving technologies, including hardware/software fine-grained power management read more
Saves power up to 65% over previous GC Cores using intelligent DVFS and incremental low power architectural enhancements
Proprietary Vega lossless compression that reduces on-chip bandwidth by an average of 3.2:1 and streamlines the graphics subsystem including the GPU,
composition co-processsor (CPC), interconnect, and memory and display subsystems
ClearView image quality – Life-like rendering with high definition detail, MSAA, and high dynamic range (HDR) color processing
Large display rendering – Up to 4K/8K screen resolution including multi-screen support
Minimal system loading — Bandwidth reduction through enhanced memory design, compression, and reduced dependency on CPU
Large display rendering – Up to 4K/8K screen resolution including multi-screen support