RE: BIRD #61 -- Characterization of Receivers

I should have posted this earlier. It was the basis for the conclusion that
we could improve the "time at threshold" receiver model that has been used
for over 30 years. I used HSPICE and a generic CMOS receiver model as a test
case. Then I ported the results into MathCad to view in 3D. I could post the
HSPICE file if you like, but its pretty simple. I wonder how many buffers
out there are this predictable. If many are, I think BIRD #61 is a
substantial improvement for what exists over "time to threshold" timing..

Richard Mellitz,
Intel

 <<input_model_Char1.pdf>>