docs: add English README translation

Author: LessUp

README.en.md

@@ -0,0 +1,59 @@
+# SGEMM Optimization: From Naive to Tensor Core
+
+[![License: MIT](https://img.shields.io/badge/License-MIT-blue.svg)](LICENSE)
+![CUDA](https://img.shields.io/badge/CUDA-11.0+-76B900?logo=nvidia&logoColor=white)
+![C++](https://img.shields.io/badge/C%2B%2B-17-00599C?logo=c%2B%2B&logoColor=white)
+
+[简体中文](README.md) | English
+
+Hand-written, progressively optimized matrix multiplication — the "Hello World" of HPC.
+
+## Performance (RTX 3060 Laptop, 1024×1024×1024)
+
+| Kernel | GFLOPS | vs cuBLAS |
+|--------|--------|-----------|
+| cuBLAS (ref) | 5727 | 100% |
+| Tensor Core (WMMA) | 2300 | 40.2% |
+| Tiled (32×32) | 753 | 13.1% |
+| Double Buffer | 701 | 12.2% |
+| Bank Conflict Free | 673 | 11.8% |
+| Naive | 604 | 10.6% |
+
+## Optimization Levels
+
+| Level | Description | Key Technique |
+|-------|-------------|---------------|
+| Naive | Basic triple loop | One thread per output element |
+| Tiled | Shared memory tiling | Data reuse, reduced global memory access |
+| Bank Conflict Free | Eliminate bank conflicts | Shared memory padding (+1) |
+| Double Buffer | Pipeline overlap | Compute/memory overlap |
+| Tensor Core | WMMA API | Hardware-accelerated matrix ops (FP16→FP32) |
+
+## Build & Run
+
+```bash
+make GPU_ARCH=sm_86   # Adjust for your GPU
+./build/sgemm_benchmark
+```
+
+## Key Optimization Techniques
+
+1. **Memory Coalescing** — Warp-aligned memory access for full bandwidth
+2. **Shared Memory Tiling** — O(N³/TILE_SIZE) global memory reduction
+3. **Bank Conflict Elimination** — +1 padding for 32x bandwidth recovery
+4. **Double Buffering** — Overlap next-tile load with current-tile compute
+5. **Tensor Core (WMMA)** — 16×16×16 hardware MMA, ~8x over CUDA Cores
+
+## Project Structure
+
+```
+├── src/kernels/           # 5 kernel implementations
+├── src/utils/             # CUDA utils, benchmark, verification
+├── src/main.cu            # Entry point
+├── tests/test_sgemm.cu    # Google Test property tests
+└── Makefile
+```
+
+## License
+
+MIT License

README.md

@@ -1,5 +1,7 @@
 # SGEMM Optimization: From Naive to Tensor Core
 
+简体中文 | [English](README.en.md)
+
 [![License: MIT](https://img.shields.io/badge/License-MIT-blue.svg)](LICENSE)
 ![CUDA](https://img.shields.io/badge/CUDA-11.0+-76B900?logo=nvidia&logoColor=white)
 ![C++](https://img.shields.io/badge/C%2B%2B-17-00599C?logo=c%2B%2B&logoColor=white)