Explore tens of thousands of sets crafted by our community.
GPU Architecture
15
Flashcards
0/15
Display Outputs
The various ports on a GPU card used to connect to displays, including HDMI, DisplayPort, DVI, and others, essential for transmitting video signals to monitors.
Render Output Units (ROPs)
Involved in the final stage of rendering 3D images, converting them into a 2D image to display on screen, and handling anti-aliasing.
Raster Operators (ROPs)
Responsible for rendering output pixels in the frame buffer; they perform final operations like blending and anti-aliasing.
Graphics Processing Clusters (GPCs)
High-level organization within NVIDIA GPUs that contains one or more Streaming Multiprocessors (SMs), which work in coordination to process large volumes of graphics data.
Unified Shader Model
A GPU design that utilizes the same shader processors to handle various types of shading operations (vertex, pixel, geometry, etc.), allowing for more flexible and efficient processing.
Direct Memory Access (DMA) Engines
Facilitates data transfer between GPU memory and system memory without burdening the CPU, improving data throughput and overall system performance.
Graphics Memory (VRAM)
Dedicated video memory used to store textures, frame buffers, and shader programs, essential for high-resolution, high-texture quality gaming and graphics.
Texture Mapping Units (TMUs)
Handle texture filtering and mapping operations, crucial for adding detail and realism to 3D graphics.
Shader Processors
Execute programmable shading code for vertex, pixel, geometry, and other shaders, which dictates the rendering of visual effects.
GPU Core Clock
The frequency at which the GPU core runs; higher frequencies can increase the processing power and thereby improve gaming and rendering performance.
PCI Express (PCIe) Interface
The standard connection interface used to slot the GPU into the motherboard, allowing communication between the two components.
Compute Units
The primary processing blocks of a GPU in AMD's architecture that contains stream processors, TMUs, and other components, contributing to parallel processing capabilities.
Memory Clock
The speed at which the GPU's memory operates, influencing transfer rates and bandwidth, directly impacting performance and frame rate.
Memory Bus
The interface between the GPU and VRAM that determines the speed at which data can be transferred, affecting overall graphics performance.
CUDA Cores (NVIDIA) / Stream Processors (AMD)
Responsible for executing instructions in parallel, which enhances the processing of graphics and compute tasks.
© Hypatia.Tech. 2024 All rights reserved.
