[hardboiled]

Speed is not the only computational constraint - the tremendous memory required for very large data sets is often a severe limitation especially within R (and I actually enjoy R).

This is Julia's native, in-language approach (no delegating to C/C++): https://juliadb.org/