Static 2$\times$superscalar - problems

New hazards:

• While Integer/FP split is simple for the HW, get CPI of 0.5 only for programs with exactly 50% FP operations

Data and control hazards cause more stalls:

• Any hazard in an inst$^{\rm n}$ package blocks second issue

Hardware complexity:

• If more inst$^{\rm n}$s issue at same time, greater difficulty of decode and issue

• Even in 2$\times$, ID must examine 2 opcodes, 6 register specifiers, & decide if 1 or 2 inst$^{\rm n}$s can issue

(H&P pp.279)