benchmarks

some random benchmarks for fun
CPU specs: Ryzen 7 260, 8 core

• L1i 32 kb per core
• L1d 32 kb per core
• L2 1 mb per core
• l3 16 mb shared

plans

• benchrunner which can output some proper statistics for comparison
• prefetcher stuff with contiguous memory access graphic latency vs different working sizes, compare against linked list pointer chasing
• array of structures vs structure of arrays
• simd/vectorisation falls into SOA kinda
• sorting a list to help with branch prediction
• false sharing messing with concurrency
• maybe LTO
• random number generator vs inbuilt