ClockTune
Jeng-Liang Tsai, Tsung-Hao Chen, and Charlie Chung-Ping Chen
As the feature size keeps shrinking and clock frequency increasing, clock design
has come to play a crucial role in determining the chip performance and
facilitating timing and design convergence. First, clock skew directly affects
chip performance in close to a one-to-one ratio since it has to be counted as
cycle time penalty. Second, incremental clock tree adjustment enables fast
design convergence by avoiding the potentially divergent design iterations.
Since designs are subjected to change on a daily basis, the clock trees need to
be incrementally adjusted accordingly with minimum changes to ensure acceptable
clock skew. Third, since interconnect delay dominates over gate delay, timing
plans often cannot be met due to some physical effects. Recently, useful-skew
concepts have also been widely proposed to speed up timing convergence in order
to compensate for the timing uncertainties resulting from physical layout. From
the above analysis, it is crucial to develop clock tuning algorithms that can
balance clock skew with minimum adjustments. Buffer-insertion/sizing and wire-sizing
techniques are suitable for clock tuning since they involve minimum routing
modifications.
ClockTune first takes the input netlist and generates the initial clock
routing using the BB+DME algorithm. It then calculates the feasible
delay and capacitance load information of each node in a bottom-up fashion.
After the desired delay and capacitance load of the root node is chosen, an
embedding with target delay and power consumption is selected in a top-down fashion.
PC (Windows based) or SUN
ClockTune v1.2 binary and inputs for Windows
ClockTune v1.0 binary and inputs for Windows
ClockTune v1.0 binary and inputs for UNIX
ClockTune v0.9 binary and inputs for Windows
ClockTune v0.9 binary and inputs for UNIX
Input files:
This version includes simultaneous buffer-insertion/sizing and
wire-sizing capability. The ClockTune v1.2 screen shot for r1 is as below.

Below is a subset of the negative-polarity DC region of r1.
This version includes the wire-sizing capability. The ClockTune v1.0
screen shot for r1 is as below.
This beta version implements the BB+DME algorithm. The program can either
randomly generate sink nodes or read an input file, construct the clock tree
topology, and output the result in a user specified postscript file. The next
version will allow users to nevigate the clock tree and tune the wire widths.
Following is the result from r3 using Elmore delay model.
If you have any technical problem, please email the author Jeng-Liang
Tsai (jltsai@cae.wisc.edu) of this tool. Please also report any bug you
find to the author. Thanks very much.
Ran-Song Tsay, "Exact Zero Skew," IEEE Int. Conference on Computer-Aided
Design (ICCAD-91), pp. 336- 339, 1991. Nov. 1991.
T.-H. Chao, Y.-C. Hsu, J.-M. Ho, K. D. Boese, and A. B. Kahng, "Zero
Skew Clock Routing with Minimum Wirelength," IEEE Trans. Circuits Syst.
-II, pp. 799- 814, 1992.
T.-H. Chen, C. Luk, H. Kim, and C. C.-P. Chen, "INDUCTWISE: Inductance-Wise
Interconnect Simulator and Extractor," 2002 IEEE/ACM international conference on
Computer-Aided Design.