Question: On modern processors, why does the floating-point add operation take longer than the integer add operation (assume the same number of bytes in each data


On modern processors, why does the floating-point add operation take longer than the integer add operation (assume the same number of bytes in each data type)? As an example, see figure 5.12 on page 523 which shows the latency of the Intel Core i7 Haswell microprocessor 4 Integer Floating point Operation Latency Issue Capacity Latency Issue Capacity Addition 1 1 3 1 1 Multiplication 3 1 1 5 1 2 Division 3-30 3-30 1 3-15 3-15 1 Figure 5.12 Latency, issue time, and capacity characteristics of reference machine operations. Latency indicates the total number of clock cycles required to perform the actual operations, while issue time indicates the minimum number of cycles between two independent operations. The capacity indicates how many of these operations can be issued simultaneously. The times for division depend on the data values
Step by Step Solution
There are 3 Steps involved in it
Get step-by-step solutions from verified subject matter experts
