Slide5.PNG

PRODUCTS

Python framework enabling RF PCELL design in CMOS

rfcell-symbol.png
EDA: RFCell

Key features

Python framework developed by Multifractal

Enables rapid OA RF PCELL development

Code in “1D wire-frame” mode – framework generates 3D

Generate complex structures with ~100 lines of code

DRC clean and process agnostic: rules loaded from tech file

Shortens development time from months to weeks

Si-proven in 28 CMOS

Integrates with Synopsys PyCell, Cadence & Keysight ADS AEL design flows

Any conceivable, fully parameterized RF CMOS layout can be defined using high-level abstracts. The coder can think in 1D wire-frame mode and focus on the RF design problem at hand. The framework then generates 3D details in an optimal and consistent way.

Both simulation layouts (Keysight ADS) and tapeout layouts (Synopsys/Cadence) are generated from the same underlying code; but different simplifications are applied, as defined by the coder/RFCell user. This allows for accurate and consistent EM simulations and eliminates the need for export/import of layouts between EDA tools.

trans_simp_0.png

Tape-out mode

trans_simp_3.png

Simulation-mode

Generate RF PCELLS easily

Simulate & tapeout the same design

RFCell_code.png

With less than 100 lines of code, complex fully parameterized structures such as the 2 layer transformer above are easily generated (it has many hidden features: in/out rotation, sandwiching of layers, etc.). That is: without the loss of design flexibility & generality.

t_line_bend_simp_1.png

All in  Python

The RFCell framework reads in process tech files to generate DRC compliant, tapeout-ready structures. Although this does not completely alleviate the burden of DRC-aware design by the coder/user, it does make the task significantly easier allowing for more effort and focus to go into the RF design.

hold_cell_simp_0_zoom1.png
hold_cell_simp_0.png

DRC clean