Behavioral Modeling of Clock/Data Recovery
May 1, 2014
Clock/Data recovery (CDR) is a tricky logic to implement correctly. To verify the clock/data recovery logic implemented in designs, the corresponding verification infrastructure needs to be modeled correctly.
This presentation aims to present the various issues faced for modeling CDR behaviorally along with their solutions.
The presentation starts off by explaining the need for CDR in the context of serial interfaces. It then explains how CDR would be done in an ideal scenario. Real issues like Jitter and PPM/Drift are then introduced along with techniques for handling these effects. Finally, a flow chart explains how an accurate behavioral model can be constructed.
Related Semiconductor IP
- nQrux® Root of Trust IP
- AXI to UCIe Bridge IP
- UCIe 2.x Controller IP
- SWI3S (SoundWire I3S Interface) Peripheral Controller Core IP
- OpenTitan-based RISC-V Secure Element
Related Blogs
- UEC-LLR: The Future of Loss Recovery in Ethernet for AI and HPC
- TSMC Yields Recovery!
- In search of recovery
- TSMC UMC Lead Semiconductor Recovery - Record Year in 2010
Latest Blogs
- From Bug Hunting to Engineering Methodology: Lessons from AI-Driven Development
- Manufacturing Intelligence: Turning EDA Data into Trusted Action
- Execute-in-Place: Getting More from Embedded NVM
- Rethinking RTL flows with AI-driven hybrid formal verification
- Arm and NVIDIA: Building the trusted compute foundation for the agentic AI era