to get .lib timing files for NLDM and CCS charac methods comparison (28nm ideally).

[pierre13]

Hello,

I am looking for a NLDM .lib file and a CCS .lib file (timing), let's say, if possible in 28nm. I am interested in getting the differences between the 2 characterization methods, reading the .lib files. I plan to consider a "simple" cell, such as an Inverter and consider the .lib syntaxes.

I don't have any solvnet account, and I don't have a corporate email (just gmail, yahoo...), so I cannot create a new solvnet account !! I think we can get some .lib file extract from there.

Does anybody can help me ??

Thanks a lot !!

 

[graphene]

what is NLDM ?? and .lib files are not available open and cannot be shared.... .
if u can tell wat exactly you want to study also mentioning the tasks u planned I can try to help u...

 

[artmalik]

NLDM : Non Linear Delay Model
CCS : Current composite Source modelling.

CCS modelling involves the current profile which the final driver of the design while nldm is older method of representing the delay of the cells.

in CCS the current profile is capture during a rise/fall delay of the cell. The output driver of the cell is represented as a current source.

The simulation is the SAME in the generation of NLDM and CCS. The representation is different. so in short there is NO script to convert NLDM model into CCS. Also there are standard tools which do that NCX(Synopsys), ALtos(Cadence) and SiliconSmart( synopsys formerly Magma).

CCS is more accurate and is used in the all the technologies startig 45nm and beyond.... we have been using CCS since 65nm....

what is NLDM ?? and .lib files are not available open and cannot be shared.... .
if u can tell wat exactly you want to study also mentioning the tasks u planned I can try to help u...

 

[rca]

to complete the artmalik informations, the NLDM is only a table, and the interpolation is linear between each point define inside this table.
In CCS, the model describe the wave form of the rise/fall delay (and other), then the calculated value is more accurate than the NLDM.

There is a third model:ECSM, Effective Current Source Model
is based on current-source modeling
consists of voltage samples as a function of time

 

[edafreek]

to complete the artmalik informations, the NLDM is only a table, and the interpolation is linear between each point define inside this table.
In CCS, the model describe the wave form of the rise/fall delay (and other), then the calculated value is more accurate than the NLDM.

There is a third model:ECSM, Effective Current Source Model
is based on current-source modeling
consists of voltage samples as a function of time

@rca so it means that we characterize both NLDM and ECSM in same way and the only difference is lies in sampling its output voltage waveform instead of only one value ??

 

[rca]

Yes, the curve granility is better with CCS & ECSM model versus NLDM.