Fig 1.
A typical quasi delay insensitive (QDI) asynchronous circuit stage.
Fig 2.
Timing behavior of indicating circuits with respect to: (a) RTZ; and (b) RTO handshaking.
Table 1.
Truth table of QDI QMR majority voter corresponding to RTZ handshaking.
Fig 3.
Safely decomposed DIMS QMR majority voter corresponding to RTZ handshaking.
The OR gates shown in red should be replaced by AND gates to obtain the RTO equivalent circuit.
Fig 4.
Dysart’s synchronous QMR majority voter.
Fig 5.
QDI asynchronous QMR majority voter corresponding to RTZ handshaking, realized based on Dysart’s synchronous QMR majority voter logic.
The gates highlighted in red should be replaced by their respective gate duals to obtain the RTO equivalent circuit.
Fig 6.
QDI QMR realization of a synchronous QMR majority voter [14], corresponding to RTZ handshaking.
The gates highlighted in red, blue and pink should be replaced by their respective gate duals to obtain the RTO equivalent.
Fig 7.
Proposed QDI QMR majority voter corresponding to RTZ handshaking.
To obtain the RTO equivalent circuit, the AND4, OR4 and OR2 gates highlighted in red should be replaced by OR4, AND4 and AND2 gates respectively.
Fig 8.
Screenshot of portion of the simulation waveform of an example QDI QMR circuit incorporating P-QMV, based on RTZ handshaking.
Fig 9.
Screenshot of portion of the simulation waveform of an example QDI QMR circuit incorporating P-QMV, based on RTO handshaking.
Table 2.
Cycle time, silicon area, and averaged (total) power dissipation of various QDI QMR circuits incorporating different QMR majority voters estimated using a 32/28nm CMOS process.
Fig 10.
Normalized figure-of-merits of QDI QMR circuits employing different QMR majority voters corresponding to RTZ and RTO handshaking: (a) Normalized PCTP plots; and (b) Normalized ECTP plots.