GitHub

intr_vector_reduce.ll

%reduce_add = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> <i32 1, i32 2, i32 3, i32 4>) %reduce_add_poison = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32 ...
GitHub

intr_fp_vector_reduce.ll

%fadd = call float @llvm.vector.reduce.fadd.v4f32(float 1.0, <4 x float> <float 2.0, float 3.0, float 4.0, float 5.0>) %fadd_poison_acc = call float @llvm.vector ...
IEEE

Auto-Vectorization Optimization for VLIW DSP Architecture Based on LLVM

Auto-vectorization is an important technique for exploiting data-level parallelism. However, existing approaches are primarily designed for SIMD-based general-purpose processors and are not ...
IEEE

Less is More: Efficient Image Vectorization with Adaptive Parameterization

Abstract: Image vectorization aims to convert raster images to vector ones, allowing for easy scaling and editing. Existing works mainly rely on preset parameters (i.e., a fixed number of paths and ...