sm75

FlashAttention v1 forward pass in CUDA for NVIDIA Turing (SM75)

Compiler MVP that detects Transformer fusion patterns, generates optimized CUDA kernels with WMMA Tensor Cores, and executes them on real GPU hardware — 10.5 TFLOPs on RTX 2070, correctness validated against PyTorch.

Backward pass CUDA kernels for fused GEMM+Bias+GeLU on SM75 (Turing). Float4 vectorized WMMA kernels validated against PyTorch autograd. 3.1x faster than autograd at M=1024. 24/24 tests passing.

Fork of FlashQLA-SM70-SM75 with the SM75 import/build tweaks used by qwen3-8-27b-dual-2080ti-vllm (branch 2080ti_dual_qwen38-27B)

Instruction-level Tensor Core micro-kernel engineering on SM75 (RTX 2070), including PTX manual emission, shared-memory staging, and hardware-aware auto-tuning.

Fork of vLLM carrying the SM75 / Qwen3.8 changes used by qwen3-8-27b-dual-2080ti-vllm (branch 2080ti_dual_qwen38-27B); upstream main is kept as the 'upstream' branch

Reproducible FreeToken MoE expert-offload compatibility lab and benchmarks for NVIDIA Turing sm_75.

Qwen3.8-27B 双 RTX 2080 Ti 22GB + NVLink 的 vLLM 部署（上游 main 路线）：SM75 移植、MTP 下保住 FULL cudagraph（单并发解码 54.1→37.5 ms/步）、256K/500K 上下文与 128K tiered offload profile