Cadence ENCOUNTER LIBRARY CHARACTERIZER Datasheet - page 5
5
www.cadence.com
ENCOUNTER LIBRARY CHARACTERIZER
advanced statistical ECSMs to determine
the probability of timing failure over the
full range of process variation and can
account for the variability of the process
parameters in a single run. This reduces
both pessimism and guardbanding, which
decreases area and power consumption,
improves chip performance by as much
as 20%, and allows designers to explore
the potential tradeoff between parametric
yield and clock speed.
Advanced features include:
• Si2-approved statistical ECSM gener-
ation for greater accuracy
• Sensitivity-based (delay per parameter
change)
• Nominal + delay due to changes in
parameters
• Within-the-die, die-to-die, and random
variation modeling support
• Support for Encounter Timing System,
Encounter Digital Implementation
System, and third-party tools
STATISTICAL LEAKAGE POWER
MODELING
At smaller process geometries, leakage
power begins to dominate the power
consumed by CMOS devices. Therefore,
accurately modeling device leakage is
critical to achieving an efficient low-power
solution. Existing leakage power analysis
techniques use a pessimistic worst-case
leakage scenario that has a very small
likelihood of occurring in real silicon. Since
cell leakage has an exponential response
to process variation—meaning that a
small change in process variation causes
a major shift in transistor leakage—the
probability distribution becomes skewed
such that the extreme worst case has a
very small chance of occurring (Figure
3). This leads to aggressive overdesign,
overcompensation for IR drop, and
potentially unnecessary architectural
changes.
The solution is to model the leakage
power as a statistical probability to avoid
designing to the worst-case limit. This
enables designers to target a smaller,
more reasonable leakage number, which
can be as much as 40% smaller than that
from traditional worst-case leakage power
analysis. Encounter Library Characterizer
GXL is foundry-certified for leakage power
characterization using the statistical ECSM
format:
• State-dependent exponential leakage
power variation with respect to process
parameters
• Die-to-die, within-the-die (spatially
correlated), and random variation
PACKAGING
Encounter Library Characterizer is
available in XL and GXL base licenses.
PLATFORMS
• Linux (32-bit, 64-bit)
• Solaris (64-bit)
• AIX (64-bit)
• Sol86 (64-bit)
STANDARD INTERFACE
SUPPORT
• Inputs: SPICE subckt, SPICE models,
.lib, setup_file (for slew, load, operating
condition information), and statistical
config file (for parameter variation)
• Outputs: .lib + ECSM timing, noise,
power, and statistical formats, and CCS
• Simulators Supported: Cadence Spectre,
Synopsys HSpice, Mentor Eldo
CADENCE SERVICES AND
SUPPORT
• Cadence application engineers can
answer your technical questions by
telephone, email, or Internet—they can
also provide technical assistance and
custom training
• Cadence certified instructors teach
more than 70 courses and bring
their real-world experience into the
classroom
• More than 25 Internet Learning Series
(iLS) online courses allow you the
flexibility of training at your own
computer via the Internet
• Cadence Online Support gives you
24x7 online access to a knowledgebase
of the latest solutions, technical
documentation, software downloads,
and more
Total leakage PDF
Encounter
Leakage (nW)
1.20E-04
1.00E-04
8.00E-05
6.00E-05
4.00E-05
2.00E-05
0.00E+00
0
10000
20000
30000
40000
50000
60000
-2.00E-05
Spice Monte Carlo
Figure 3: Statistical leakage power model: Encounter Power System vs. Monte Carlo SPICE