Zapcc Still Aiming For "Super Fast" Compiler Performance
It's been a while since last covering Zapcc as a new, super-fast C/C++ compiler yet it has evolved and now the latest beta is reporting to show even more impressive performance gains.
Zapcc is a compiler based on LLVM/Clang that has routinely strived for maximum performance not just for compile-time performance but also the resulting performance of the compiled binaries.
With the latest claims of the company in their newest announcement includes Zapcc performance stating:
Zapcc is a compiler based on LLVM/Clang that has routinely strived for maximum performance not just for compile-time performance but also the resulting performance of the compiled binaries.
With the latest claims of the company in their newest announcement includes Zapcc performance stating:
Acceleration numbers of Zapcc vary by the project. Some examples:Their reported numbers are certainly promising, it will be fun and interesting to see where Zapcc leads into the future. See the previous links for more details.
Zapcc accelerates Webkit full build by x6.81 compared to Clang
Paraview full build is accelerated by x2.13
Boost.Math Examples full build is x5.11 faster
And even Clang/LLVM is built x2.07 faster using Zapcc
Incremental build times are typically x5-x10 faster (and we have seen much more, up to x100), depending on exact scenario and which files were changed. Such speedups dramatically increase the productivity of developers (and their joy ...).
Zapcc uses in-memory compilation cache in a client-server architecture, remembering all compilation information between runs. It does not use distributed network servers. This means you can accelerate your build times without additional hardware, and at a much lower energy footprint.
11 Comments