Skip to main content

Crate tpt_gpu_kernel_optimizer

Crate tpt_gpu_kernel_optimizer 

Source
Expand description

TPT GPU kernel parameter optimizer.

Three-phase search strategy:

  1. Grid search — exhaustive sweep of the full parameter space
  2. Hill-climbing — greedy local improvement from the best grid result
  3. AI-guided search — LLM suggests next candidates based on history

Re-exports§

pub use real_evaluator::generate_milestone_json;
pub use real_evaluator::generate_milestone_report;
pub use real_evaluator::optimize_all_gemm_problems;
pub use real_evaluator::optimize_gemm_problem;
pub use real_evaluator::standard_gemm_problems;
pub use real_evaluator::GemmOptResult;
pub use real_evaluator::GemmProblemConfig;
pub use real_evaluator::RealGemmEvaluator;
pub use attention_eval::generate_attention_milestone_json;
pub use attention_eval::generate_attention_milestone_report;
pub use attention_eval::optimize_all_attention_problems;
pub use attention_eval::optimize_attention_problem;
pub use attention_eval::standard_attention_problems;
pub use attention_eval::AttentionOptResult;
pub use attention_eval::AttentionProblemConfig;
pub use attention_eval::RealAttentionEvaluator;
pub use fused_eval::fused_gemm_problems;
pub use fused_eval::generate_beat_cublas_json;
pub use fused_eval::generate_beat_cublas_report;
pub use fused_eval::optimize_fused_problem;
pub use fused_eval::run_beat_cublas_campaign;
pub use fused_eval::FusedGemmEvaluator;
pub use fused_eval::FusedGemmOptResult;
pub use fused_eval::FusedGemmProblem;
pub use conv2d_eval::generate_conv2d_milestone_json;
pub use conv2d_eval::generate_conv2d_milestone_report;
pub use conv2d_eval::optimize_all_conv2d_problems;
pub use conv2d_eval::optimize_conv2d_problem;
pub use conv2d_eval::standard_conv2d_problems;
pub use conv2d_eval::Conv2dOptResult;
pub use conv2d_eval::Conv2dProblemConfig;
pub use conv2d_eval::RealConv2dEvaluator;
pub use normalization_eval::generate_normalization_milestone_json;
pub use normalization_eval::generate_normalization_milestone_report;
pub use normalization_eval::optimize_all_normalization_problems;
pub use normalization_eval::optimize_normalization_problem;
pub use normalization_eval::standard_normalization_problems;
pub use normalization_eval::NormalizationOptResult;
pub use normalization_eval::NormalizationProblemConfig;
pub use normalization_eval::RealNormalizationEvaluator;
pub use vector_add_eval::generate_vector_add_milestone_json;
pub use vector_add_eval::generate_vector_add_milestone_report;
pub use vector_add_eval::optimize_all_vector_add_problems;
pub use vector_add_eval::optimize_vector_add_problem;
pub use vector_add_eval::standard_vector_add_problems;
pub use vector_add_eval::RealVectorAddEvaluator;
pub use vector_add_eval::VectorAddOptResult;
pub use vector_add_eval::VectorAddProblemConfig;
pub use kernel_category::KernelCategory;

Modules§

attention_eval
Attention evaluator for the optimizer loop.
conv2d_eval
Conv2D evaluator for the optimizer loop.
fused_eval
Fused GEMM evaluator — models GEMM+bias+activation fusion vs cuBLAS pipeline.
kernel_category
Kernel category classification and kernel-specific parameter spaces.
normalization_eval
Normalization evaluator for the optimizer loop.
real_evaluator
Real-hardware evaluators for the optimizer loop.
vector_add_eval
VectorAdd evaluator for the optimizer loop.

Structs§

OptResult
Optimization result for a single parameter configuration.
ParamDim
A single tuning dimension: name + allowed discrete values.
ParamSpace
The full discrete parameter space.
SimulatedEvaluator
Synthetic evaluator — no hardware required.
TuningParams
A concrete point in the tuning parameter space.

Traits§

KernelEvaluator

Functions§

ai_guided_search
LLM-guided search: ask the AI for the next candidate, evaluate, repeat.
grid_search
Evaluate every point in the space; return sorted results (best first).
hill_climb
Greedy hill-climb from start; stops when no neighbor improves the score.

Type Aliases§

ParamVal