Comment by n0n0n4t0r

How a friend of mine created a super fast csv parser by playing with branchless programming and SIMD to avoid the CPU serialized computation bottleneck