Online HelpPC 2.10      Quick Reference Utility     Copyright 1991 David Jurgens
 
                     Instruction Clock Cycle Calculation                     
                                                                                                                                                                          Some instructions require additional clock cycles due to a "Next                 Instruction Component" identified by a "+m" in the instruction                   clock cycle listings.  This is due to the prefetch queue being                   purge on a control transfers.   Below is the general rule for                    calculating "m":                                                                                                                                                                                                                                   88/86 not applicable                                                             286  "m" is the number of bytes in the next instruction                          386  "m" is the number of components in the next instruction                             (the instruction coding (each byte), plus the data and                           the displacement are all considered components)                                                                                                                                                                                   
                8088/8086  Effective Address (EA) Calculation                
                                                                                
                   Description                            Clock Cycles          
                                                                                         Displacement                                            6                        Base or Index (BX,BP,SI,DI)                             5                        Displacement+(Base or Index)                            9                        Base+Index (BP+DI,BX+SI)                                7                        Base+Index (BP+SI,BX+DI)                                8                        Base+Index+Displacement (BP+DI,BX+SI)                  11                        Base+Index+Displacement (BP+SI+disp,BX+DI+disp)        12                                                                                                                                                                                          - add 4 cycles for word operands at odd addresses                                - add 2 cycles for segment override                                              - 80188/80186 timings differ from those of the 8088/8086/80286                                                                                                                                                                            
        Task State Calculation                                                  
                                                                                         "TS" is defined as switching from VM/486 or 80286 TSS to one of                  the following:                                                                                                                                                                    ┌───────────────────────────────────────┐                                        │               New Task                │                                        ├───────┬───────┬───────┬───────┬───────┤                        ┌───────────────┤486 TSS│486 TSS│386 TSS│386 TSS│286 TSS│                        │   Old Task    │ (VM=0)│ (VM=1)│ (VM=0)│ (VM=1)│       │                        └───────────────┼───────┼───────┼───────┼───────┼───────┤                        386 TSS (VM=0)  │       │       │  309  │  226  │  282  │                                        ├───────┼───────┼───────┼───────┼───────┤                        386 TSS (VM=1)  │       │       │  314  │  231  │  287  │                                        ├───────┼───────┼───────┼───────┼───────┤                        386 CPU/286 TSS │       │       │  307  │  224  │  280  │                                        ├───────┼───────┼───────┼───────┼───────┤                        486 CPU/286 TSS │  199  │  177  │       │       │  180  │                                        └───────┴───────┴───────┴───────┴───────┘                                                                                                                                                                                 
        Miscellaneous                                                           
                                                                                         - all timings are for best case and do not take into account wait                  states, instruction alignment, the state of the prefetch queue,                  DMA refresh cycles, cache hits/misses or exception processing.                 - to convert clocks to nanoseconds divide one microsecond by the                   processor speed in MegaHertz:                                                                                                                          
                      (1000MHz/(n MHz)) = X nanoseconds                      
                                                                                                                                                                          - see   8086 Architecture                                                                                                                                                                                                                   
[helppc.bosbyte.nl]           Instruction Timing              Home  Back  Topics