Which is easier depends on a lot of things but basically IMHO
SKILL pcells are easier if:
= You are dealing in legacy technology - 180nm and above
= You have tons of legacy SKILL infrastructure and existing devices
= You have a CAD team of crusty SKILL coders who are change-adverse
= Cadence is willing to develop them for you for free (they weren't for our company)
pyCells are easier if:
= You are moving to a 40nm and below technology - some effects are nigh impossible to code in SKILL
= Your CAD team wants to learn a modern, useful langauge
= You don't have legacy devices or are willing to restart - we did this when we moved to OA. It was worthwhile giving up all the .cdb based cruft.
Comment - This all assumes you are using CiraNova's free PyCell studio which really makes doing pyCells easy and interactive.
No doubt, Cadence is scared to join iPDK, because PDK is bread and butter biz of CADENCE. and it would be foolish of them to join iPDK.
By the as a end user which one is easy to use PCELL from SKILL or PyCell?
It's sad to see Cadence backing into the proprietary SKILL language corner when OpenAccess, the database they pioneered, easily interfaces to more capable open languages like python and tcl for PDK development. TSMC and others have proved that a single open PDK can serve the industry and even work in Virtuoso.
At least they were honest four years ago when people were asking them to open up PDKs.
"That's the last closed citadel in the EDA world for interoperability."
"Bluntly, that's not going to happen," Vucurevich replied. "We're not going to open from a competitive point of view and say here's a free ride for coming in and trying to take over accounts."
Blog Doing Math in FPGAs Tom Burke 24 comments For a recent project, I explored doing "real" (that is, non-integer) math on a Spartan 3 FPGA. FPGAs, by their nature, do integer math. That is, there's no floating-point ...