A performance driven macro-cell placement algorithm

Tong Gao, Pravin M. Vaidya, C. L. Liu

Design Automation Conference · 1992 · 38 citations · 23 references

Concepts

Abstract

The authors present a new performance driven macro-cell placement algorithm. Placement of modules is guided by a set of upper- and lower-bounds on the net wire lengths. A convex programming algorithm is used to compute a set of upper-bounds on the net wire lengths which will ensure that timing requirements between input and output signals are satisfied. A set of lower-bounds is also computed to control signal skews at intermediate points of the circuit. Artificial nets are introduced between all pairs of modules. Lower-bounds on the lengths of the artificial nets are computed to avoid module overlaps in the placement. A modified min-cut placement algorithm is then used to generate a placement that satisfies the upper- and lower-bounds. An iterative procedure is used to modify the set of upper- and lower-bounds to improve the quality of the placement result. Experimental results on eight test examples are included. >

References

23