Open-source EDA Γ AI for chip design β graph-native, local-first tools for STA, GPU-parallelism, and hardware-security analysis.
- π Currently building Synthesis_project -- a research-prototype EDA suite that treats post-synthesis netlists as graphs and layers analysis (STA-lite, parallelism profiling, hardware-security audit) on top.
- π― Headline result: 14Γ theoretical parallel speedup on a 1278-node 16Γ16 multiplier (GL0AM-regime, NVIDIA-inspired analysis).
- π v0.1.0 tagged Β· 52 tests Β· CI on Python 3.10 / 3.11 / 3.12 Β· live Streamlit demo.
- π± Learning: GNNs for circuit analysis, local-LLM agents (Ollama / Llama-3 / Phi-3), GPU-accelerated logic simulation.
- π« Reach me: almallikarjun55@gmail.com
EDA Β· RTL synthesis Β· static timing analysis Β· hardware security Β· GPU-accelerated logic simulation Β· graph algorithms Β· AI for chip design
Currently building toward an Ollama-backed local LLM agent + a GNN criticality-prediction experiment.
Open to collaboration on AI-for-EDA research.