Benchmarks
Test Environment
Hardware Configuration
| Component | Specification |
| Computer Model | Lenovo ThinkPad P15v Gen 3 |
| CPU | 12th Gen Intel Core i7-12800H (20 logical, 14 physical cores) |
| Base Clock | 2.80 GHz |
| Turbo Clock | 4.80 GHz |
| L1 Data Cache | 48 KiB (×6 P-cores) + 32 KiB (×8 E-cores) |
| L1 Instruction Cache | 32 KiB (×6 P-cores) + 64 KiB (×2 E-core clusters) |
| L2 Unified Cache | 1.25 MiB (×6 P-cores) + 2 MiB (×2 E-core clusters) |
| L3 Unified Cache | 24 MiB (×1 shared) |
| RAM | DDR4-3200 (32GB) |
| GPU | NVIDIA RTX A2000 4GB GDDR6 |
Software Configuration
| Platform | OS | Benchmark Framework | C++ Compiler | nfx-cpu Version |
| Linux | LMDE 7 | Google Benchmark v1.9.4 | GCC 14.2.0-x64 | v0.1.0 |
| Windows | Windows 10 | Google Benchmark v1.9.4 | MinGW GCC 14.2.0-x64 | v0.1.0 |
| Windows | Windows 10 | Google Benchmark v1.9.4 | Clang-GNU-CLI 19.1.5-x64 | v0.1.0 |
| Windows | Windows 10 | Google Benchmark v1.9.4 | Clang-MSVC-CLI 19.1.5-x64 | v0.1.0 |
| Windows | Windows 10 | Google Benchmark v1.9.4 | MSVC 19.44.35217.0-x64 | v0.1.0 |
Performance Results
CPU Detection Overhead
| Operation | Linux GCC | Linux Clang | Windows MinGW GCC | Windows Clang-GNU-CLI | Windows Clang-MSVC-CLI | Windows MSVC |
| BM_Sse42Detection_Cached | 0.217 ns | 0.235 ns | 0.299 ns | 0.240 ns | 0.517 ns | 0.726 ns |
| BM_AvxDetection_Cached | 0.216 ns | 0.238 ns | 0.315 ns | 0.238 ns | 0.513 ns | 0.721 ns |
| BM_Avx2Detection_Cached | 0.217 ns | 0.235 ns | 0.313 ns | 0.274 ns | 0.517 ns | 0.728 ns |
| BM_RawCPUID | 324 ns | 321 ns | 438 ns | 647 ns | 355 ns | 335 ns |
Benchmarks executed on November 25, 2025
1.15.0