If no, how can we calculate double-precision performance for a CPU? Especially nowadays cpus are multi-threaded, the formula really confused me.
I have done researched on the internet , and the most convincing formula is
sockets * (cores per socket) * (number of clock cycles per second) * (number of floating point operations per cycle).
E.g. for the i7-7700 , the GFLOPs will be 1 socket * 4 physical cores * 4 cycle/s(all 4 cores enabled) * 16(flopc)
I have done researched on the internet , and the most convincing formula is
sockets * (cores per socket) * (number of clock cycles per second) * (number of floating point operations per cycle).
E.g. for the i7-7700 , the GFLOPs will be 1 socket * 4 physical cores * 4 cycle/s(all 4 cores enabled) * 16(flopc)