My code is based on ccminer, and alexis78 is one of the contributors. The reason why the file is renamed is that it has been optimized to be faster than in the alexis fork. Alexis improved echo by removing some of the sharemem bank conflicts in AES. He solved this bymoving some of the tables to the global memory (level1 cache).
I increased the shared mem buffer by having several copies of the the aes table and removed all of the sharemem conflicts. this resulted in a +30% faster echo. The alexis echo function is based on my work back in 2015. Back then I  precualculated the P[] table to save some AES rounds among other improvements.