Physical Lint -- Better RTL Quality Improves Design Convergence
Mark Baker, Atrenta
EETimes (6/8/2015 07:09 PM EDT)
Higher quality of RTL delivered to the physical design team will have an impact on tge physical design team's ability to achieve design closure.
The quality of RTL delivered to the physical design team will have a major impact on design convergence, or more explicitly, the ability of the physical design team to achieve design closure.
Quality, in this case, refers to metrics used to assess the physical complexity of logic structures in the design. Design closure for completeness refers to closing timing, routability and area through physical design.
Typically, the analysis of physical feasibility (Figure 1) starts in the latter stages of physical synthesis and often carries through several cycles of physical design (place & route). Design teams will acknowledge that these iterations can be reduced if logic changes are made along critical design paths, however, back-end engineers generally lack insight into the design functionality, making changes impractical at this stage of the flow. Therefore, the investigation to narrow down and resolve the design issues often remains with the physical design team through a series of iterations that rely on floorplan changes, physical constraints, or high effort design optimization techniques. It’s not until these methods are exhausted that the issue will be passed back to the front-end designers. Once the relevant changes are made the entire cycle starts again. It is now easy to see why design schedules are so unpredictable.
To read the full article, click here
Related Semiconductor IP
- 12-bit, 400 MSPS SAR ADC - TSMC 12nm FFC
- UA Link DL IP core
- 10-bit Pipeline ADC - Tower 180 nm
- NoC Verification IP
- Simulation VIP for Ethernet UEC
Related Articles
- Physical Design Exploration of a Wire-Friendly Domain-Specific Processor for Angstrom-Era Nodes
- CRADLE: Conversational RTL Design Space Exploration with LLM-based Multi-Agent Systems
- AI in VLSI Physical Design: Opportunities and Challenges
- Multimodal Chip Physical Design Engineer Assistant
Latest Articles
- Analog Foundation Models
- Modeling and Optimizing Performance Bottlenecks for Neuromorphic Accelerators
- RISC-V Based TinyML Accelerator for Depthwise Separable Convolutions in Edge AI
- Exclude Smart in Functional Coverage
- A 0.32 mm² 100 Mb/s 223 mW ASIC in 22FDX for Joint Jammer Mitigation, Channel Estimation, and SIMO Data Detection