Repository navigation
Releases
Releases · nagarjunak-pixel/aether-compiler
AetherCompiler v1.0.0 — MLIR/Triton SSA Graph Compiler, Polyhedral H100 Autotuner & C++20 Arena Planner
AetherCompiler v1.0.0 — MLIR/Triton SSA Graph Compiler, Polyhedral H100 Autotuner & C++20 Arena Planner
Latest
Initial v1.0.0 release of AetherCompiler (aether-compiler):
- Typed SSA Tensor Graph IR & Multi-Pass Optimizer:
AlgebraicSimplify+CSE -> VerticalEpilogueAndResidualFusion -> DeadCodeElimination(17 -> 9IR nodes,2.18xHBM traffic reduction,|Δ| = 2.38e-5, strict SSA dominance & shape verified). - Native C++20 Linear-Scan Static Arena & Tiled Fused GEMM:
clang++ -O3best-fit 256B-aligned liveness allocator compressing 2-layer Transformer activations from432 MiB -> 120 MiB(3.60x,0live-buffer overlaps,100%C++20-vs-Python offset parity, fusedGEMM+Bias+SiLU |Δ| = 8.3e-7). - Polyhedral Hopper H100 SXM5 Tile & CuTe Bank-Conflict Autotuner: Eliminates
32-way -> 1-wayshared-memory bank conflicts via bijectiveCuTeXOR swizzling and selectsCTA_256x256x32_S4_XOR(411.65 TFLOP/s,31.02xspeedup over naive32x32). - Live
qwen3.5:4bHardware Cost-Model CEGIS Gate: Lifts rawPass@1from4/6(66.7%) to6/6(100.0%,+33.3%lift,0OOM/misalign leaks).