- 100% C (C++ headers), as simple as memcpy<p>- No other base64 library encode or decode faster<p>- Scalar can be faster than other SSE or ARM Neon based base64 libraries<p>- SSE faster than other SSE/AVX/AVX2! base64 library<p>- Fastest AVX2 implementation
TurboBase64 AVX2 decoding up to ~2x faster than other AVX2 libs.<p>- TurboBase64 is 3-4 times faster than other libs for short strings<p>- Fastest ARM Neon base64<p>- (2023.04) avx512 - 2x faster than avx2, faster than any other implementation<p>- Dynamic CPU detection and JIT scalar/sse/avx/avx2/avx512 switching<p>- Base64 robust error checking, optimized for long+short strings<p>- Portable library, 32/64 bits, SSE/AVX/AVX2/AVX512, ARM Neon, Power9 Altivec<p>- OS:Linux amd64, arm64, Power9, MacOs+Apple M1, s390x. Windows: Mingw, visual c++<p>- Big endian + Little endian<p>- Ready and simple to use library, no armada of files, no hassles dependencies
Neat library, upvoted.<p>Question: does this business model work well? The GPL+commercial license on request?<p>I’m hoping the answer is yes, would be great with some nuance/details/exposition, thanks!