How arjavsharma91/v4q128 works
v4q128 is a header-only C++ library for four-lane, 128-bit fixed-point SIMD arithmetic using AVX2. Users construct vectors or load AoS data into the library's SoA representation, apply arithmetic and comparisons, then store or inspect results. The graph covers the sampled storage, conversion, and arithmetic modules; the README describes the product but does not document extra capabilities or their wiring.
Main entry points: storage.hpp Vector storage · transpose.hpp Load and transpose · add_sub.hpp Add and subtract · cmp.hpp Comparisons
Components 7
Vector representation
- Vector storage
include/v4q128/core/storage.hpp
Memory conversion
- Load and transpose
include/v4q128/io/transpose.hpp
SIMD operations
- Add and subtract
include/v4q128/ops/add_sub.hpp - Comparisons
include/v4q128/ops/cmp.hpp - Multiplication
include/v4q128/ops/mul.hpp - Shift helpers
include/v4q128/ops/shifts.hpp
Other components
- Library user
Connections 11
- Library user → Vector storage: constructs vectors
include/v4q128/core/storage.hpp - Library user → Load and transpose: loads and stores
include/v4q128/io/transpose.hpp - Library user → Add and subtract: adds or subtracts
- Library user → Multiplication: multiplies
- Library user → Comparisons: compares
include/v4q128/ops/cmp.hpp - Library user → Shift helpers: uses shifts
- Load and transpose → Vector storage: creates vectors
include/v4q128/io/transpose.hpp - Add and subtract → Vector storage: returns vectors
include/v4q128/ops/add_sub.hpp - Multiplication → Vector storage: returns vectors
include/v4q128/ops/mul.hpp - Comparisons → Vector storage: reads vectors
include/v4q128/ops/cmp.hpp - Shift helpers → Vector storage: operates on vectors
include/v4q128/ops/shifts.hpp