The Small/simple C/C++ Library (SCL) provides various algorithms in a portable way, with little to no external dependencies. The library is written in C++17 and uses the MPL-2.0 License or Public Domain dedications.
The subdirectories are structured as follows:
mwc256/contains a Public Domain 256 bit Multiply-With-Carry PRNG implementation for generic CPUs (ALU). With just a multiplication and add per round, this PRNG is as simple as it can get. On 64 bit machines it is probably the fastest single-data PRNG implementation, it passes PractRand at 32TB.shishua/contains a Public Domain Shishua PRNG implementation for generic CPUs (ALU), the SSE2/3 and AVX2 instruction set variants. This is currently the fastest known PRNG implementation, it passes PractRand at 32TB.chacha/contains a Public Domain ChaCha block cipher implementation for generic CPUs (ALU), the SSE2/3 and AVX2 instruction set variants. This is probably the fastest CSPRNG implementation, it passes PractRand at 32TB.keccak/contains a Public Domain Keccak PRNG implementation for generic CPUs (ALU) and system entropy gathering code according to the "Welcome to the Entropics: Boot-Time Entropy in Embedded Devices" paper. Using Keccak allows for a 1600 bit CSPRNG entropy pool, it passes PractRand at 32TB.spline/contains a Public Domain cubic spline interpolation implementation, which can be used to create an approximation of a function given a number of knots (control points) with third-order polynomial segments connecting each pair of data points.