John D McCalpin questions

John D McCalpin answers

• Does NUMA impact memory bandwidth, or just latency?
• Unaligned access performance on Intel x86 vs AMD x86 CPUs
• Benchmarking memory copy in a single shot
• Do any CPU architectures use Metadata?
• Can a lower level cache have higher associativity and still hold inclusion?
• Multiply-add vectorization slower with AVX than with SSE
• what does STREAM memory bandwidth benchmark really measure?
• programmatically disable hardware prefetching on AMD systems
• NUMA Get Current Node/Core
• how to do mmap for cacheable PCIe BAR