Memory access patterns are usually the main reason for low performance and should be optimized first because it's a great low hanging fruit (and usually the big

Memory access patterns are usually the main reason for low performance and should be optimized first because it's a great low hanging fruit (and usually the bigger problem is latency, not bandwidth). And don't think of waiting for memory as making CPU optimizations irrelevant, but instead as an oppurtinity to hide more CPU operations in the remaining 'memory access gaps' (e.g. the CPU won't simply stop working when waiting for data to be loaded from memory, it can continue with other things that don't depend on that data).