float3

float3 is a three-component vector type used in shading languages such as HLSL. It represents a 3D vector of 32-bit single-precision floats and is commonly used to encode positions, directions, normals, or RGB colors in graphics pipelines.

A float3 contains components named x, y, and z, with alternative swizzles using r, g, and b. Many languages allow accessing and assigning via .x/.y/.z or .r/.g/.b, and swizzling enables combinations (for example, v.xy or v.yz).

Operations on float3 include vector arithmetic (addition, subtraction, scaling), dot product (dot(a,b)), cross product (cross(a,b)), length(v) for magnitude, and normalization (normalize(v)). Most shading languages provide operator overloading so expressions like a + b or a * s are valid for float3 values.

In shaders, float3 is used for 3D positions in world or view space, direction vectors, normals, or color triplets in RGB. It participates in lighting calculations, shading, and geometric transformations, and is frequently converted to or from other vector types such as float4 or matrix representations during pipeline stages.

Memory and layout: a float3 consists of 12 bytes. In practice, GPU memory layouts often apply padding to align data structures to 16-byte boundaries, especially within constant buffers or structured buffers, to satisfy alignment requirements of the hardware.

Example in HLSL: float3 position = float3(1.0, 2.0, 3.0); float3 dir = normalize(position);