CMOS Analog VLSI Design

MILIND AMBARDEKAR Self employed

Initially, I wasn’t sure what to expect from this course. Coming from an automotive background, CFD had always felt a bit like a black box beyond post-processing plots. The sections on the Navier–Stokes equations and finite volume discretization helped connect the math to what’s actually happening in the solver. Seeing how grid generation and boundary layer resolution affect results made a lot of sense, especially when thinking about under-hood airflow and thermal management in automotive applications. One area that stood out was the discussion around convergence and stability. A real challenge during the assignments was dealing with a case that simply wouldn’t converge because of poor meshing near walls. That was frustrating, but also realistic. In aerospace projects, especially around external aerodynamics and airfoil analysis, the same issues show up if y+ and turbulence modeling aren’t handled carefully. A practical takeaway was learning a basic checklist before trusting results: mesh quality, residual trends, and sensitivity to boundary conditions. That’s already been applied to a cooling flow study at work. Overall, it felt grounded in real engineering practice.

sandeep saroj

Valuable content

Sayali Shinde CR Manager

Good Course

Sateesh Kumar Yadav PhD

Needed material that would stand up to peer review, and this mostly did, even though it’s a CAD course not code. The AWM bolt assembly chapter where he constrains the lugs and then shows the tolerance stack before adding mates felt like reading a clean PR in a shared repo; you see the intent, not just clicks. I liked the aside on configurations for barrel lengths and how that mirrors feature flags in prod, though the config table example could’ve gone a bit further into naming conventions. There’s a steady comparison of legacy drawings versus parametric workflows that maps well if you live between old arch docs and modern CI, infra, obs, even k8s mental models. wasn't sold on the trigger pack segment since the dimensions stay a bit hand-wavy, but the exploded view timing was right. It’s helped settle some fuzzy calls around when to lock dimensions versus keep them flexible, which tends to bite during PR review.