ARM.SoC.Architecture

4 arm implementation the arm implementation follows a

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: lapping property of the phase 1 and phase 2 clocks ensures that there are no race conditions in the circuit. The normal timing of the datapath components in a 3-stage pipeline is illustrated in Figure 4.9 on page 87. The register read buses are dynamic and are precharged during phase 2 (here 'dynamic' means that they are sometimes undriven and retain their logic values as electrical charge; charge-retention circuits are used to give pseudo-static behaviour so that data is not lost if the clock is stopped at any point in its cycle). When phase 1 goes high, the selected registers discharge the read buses which become valid early in phase 1. One operand is passed through the barrel shifter, which also uses dynamic techniques, and the shifter output becomes valid a little later in phase 1. The ALU has input latches which are open during phase 1, allowing the operands to begin combining in the ALU as soon as they are valid, but they close at the end of phase 1 so that the phase 2 precharge does not get through to the ALU. The ALU then continues to process the operands through phase 2, producing a valid output towards the end of the phase which is latched in the destination register at the end of phase 2. Datapath timing Figure 4.8 scheme. 2-phase non-overlapping clock ARM implementation 87 Figure 4.9 ARM datapath timing (3-stage pipeline). Note how, though the data passes through the ALU input latches, these do not affect the datapath timing since they are open when valid data arrives. This property of transparent latches is exploited in many places in the design of the ARM to ensure that clocks do not slow critical signals. The minimum datapath cycle time is therefore the sum of: the register read time; the shifter delay; the ALU delay; the register write set-up time; the phase 2 to phase 1 non-overlap time. Of these, the ALU delay dominates. The ALU delay is highly variable, depending on the operation it is performing. Logical operations are relatively fast, since they involve no carry prop...
View Full Document

This document was uploaded on 10/30/2011 for the course CSE 378 380 at SUNY Buffalo.

Ask a homework question - tutors are online