Optimize program loops through unrolling, fusion, tiling, and vectorization.
npx skills add https://github.com/rainoftime/pl-skills --skill loop-optimizer
Please help me install this Agent Skill. Skill: loop-optimizer Source: https://github.com/rainoftime/pl-skills/tree/main/loop-optimizer Command: npx skills add https://github.com/rainoftime/pl-skills --skill loop-optimizer
This Skill addresses the performance bottlenecks often found in loops by applying advanced transformations to make code run faster and more efficiently.
Apply loop unrolling and vectorization to the provided C++ code snippet to enhance its performance.
High-intent search queries and answers about installing and using this skill.
Loop performance is optimized by applying transformations like vectorization and tiling to leverage CPU cache and SIMD instructions. This Skill automates unrolling, fusion, and tiling to enhance execution speed in numerical computing and compiler development scenarios.
Loop unrolling is a transformation technique that expands loop bodies to reduce overhead and enable better instruction scheduling. Combined with dataflow analysis, it tracks dependencies accurately to maximize performance in high-performance computing tasks.
Dataflow analysis and SSA construction are required dependencies for accurate dependency tracking during loop transformation. They provide the necessary control flow and data dependency information to safely apply unrolling, fusion, tiling, and vectorization.
Optimizing matrix multiplication with loop tiling involves partitioning the computation into cache-friendly blocks. This Skill applies tiling and vectorization together to leverage CPU cache and SIMD instructions, delivering significant speedup for numerical computations.
Loop fusion and unrolling can be applied to scientific simulation code where loops are dominant. This Skill transforms loops through fusion to reduce overhead and unrolling to improve instruction-level parallelism for performance enhancement.
whtooAutomate compiler backend tasks for IR, CFG, SSA, and optimization passes.
gofhirBenchmark and profile Go code using the standard testing framework and pprof.
jkitchinProvide expert guidance for JAX numerical computing with grad, jit, vmap, and pmap.
OneWave-AIReview code for security vulnerabilities, performance bottlenecks, and maintainability issues.
