উত্তর বঙ্কিমচন্দ্র নিজেই দিয়ে গেছেন।

The Century of Computational Modeling We stand at an inflection point in human capability. For the first time in modern history, we possess both the computational power and mathematical frameworks to model geometry with precision and approximate its physical behavior. This isn't merely an incremental improvement, it represents a fundamental transformation in how we understand, design, and interact with the physical world. Autodesk - 80's CAD The Paradox of Modern Engineering Walk into any contemporary engineering firm, architecture studio, or product design lab, and you'll encounter teams fluent in Computer Aided Design and Engineering software. Millions worldwide create 3D models, run analyses, and derive patterns with ease. Yet beneath this more modern tooling lies a gap. Last week, a professor at a top engineering school shared something striking to me. The course he'd just launched was the first and only elective course at the university teaching computation geometry and physics. And this isn't just a one-off. Many students graduate capable of operating this software yet the functionality behind it remains a black box. We've democratized the operation of these tools without democratizing the understanding of their inner workings. This distinction matters more than we might think. The Hidden Language of Reality NURBS - Wikipedia Beneath every surface render and every finite element analysis lies an intricate collections of algorithms: - NURBS (Non-Uniform Rational B-Splines) and now SDF (Signed Distances Fields) that mathematically define curves and surfaces with precision - Mesh generation algorithms that break continuous geometry into discrete elements for analysis - Constraint solvers that use numerical optimization to find solutions satisfying multiple simultaneous requirements - Physics engines implementing everything from Navier-Stokes equations for fluid dynamics to finite element methods for structural analysis Understanding these fundamental algorithms is not mere academic curiosity, it's what separates users from true creators. When we grasp how computational geometry and physics operate, we gain the power to model ANY problem in terms of its geometry, physics, constraints, and objectives. This unlocks the ability to iterate through thousands if not millions of possibilities, exploring solutions and innovations that would be unimaginable without this foundational knowledge. The Unrealized Potential NASA - X-59 CFD simulation Consider the scope of what becomes possible: every scientific hypothesis involving physical space, material properties, or dynamic behavior can potentially be modeled geometrically and simulated computationally. We're leaving vast territories of this potential unexplored, held back by barriers to entry that remain unnecessarily high. A biologist with revolutionary insights about protein folding mechanisms may lack the computational geometry knowledge to model their hypothesis. A materials scientist envisioning novel composite structures may be unable to bridge the gap between physical intuition and numerical analysis. An aerodynamicist with an intuition on shock-induced flow separation may lack the computational geometry knowledge to robustly parametrize and discretize complex 3D surfaces The pattern is clear: brilliant domain expertise is bottlenecked by technical barriers that shouldn't exist. The Coming Transformation Industry trends point unmistakably toward fundamental change. A new role is emerging in what I call "Design Automation Engineers." Just as hand-drafting gave way to digital CAD in the 1980s and 90s, the era of manually constructing models and running individual simulations is ending. Increasingly, engineers are automating entire workflows: generating thousands of design variants, running optimization loops overnight, and focusing their effort on higher-level problem formulation rather than repetitive execution. This shift demands a different kind of expertise, one that combines deep algorithmic understanding with domain knowledge and the ability to encode engineering judgment as computational rules. A Call to Action The path forward requires deliberate effort across multiple fronts: Reimagine engineering education: Computational geometry and physics simulation should stand alongside calculus, linear algebra, and classical physics as core competencies. Democratize the tools: Open-source geometry ad physics simulation frameworks should be as accessible and well-documented as modern programming libraries. Build interdisciplinary bridges: Create programs that deliberately combine domain expertise in biology, materials science, urban planning, and other fields with computational modeling capabilities. The most powerful innovations will emerge at these intersections. The Stakes The century ahead will belong to those who can speak to computers in their native language: the precise, mathematical description of shapes and the fundamental physics governing how those shapes move, deform, and interact in space. This literacy will be as foundational to 21st-century innovation as calculus was to the industrial revolution. By lowering barriers and fostering deeper understanding, we can transform a generation of tool operators into tool creators, people who don't just use simulation software, but extend it, adapt it, and apply it to problems its original designers never imagined. The computational modeling revolution isn't coming. It's here. The question is whether we'll harness its full potential or remain satisfied with scratching its surface.

In 1975, Steve Jobs was working at Atari, a video game company. Atari’s founder, Nolan Bushnell, asked Jobs to help design a circuit board for a new version of the game Breakout. Jobs wasn’t an engineer, so he turned to his friend Steve Wozniak, who was brilliant at circuit design. Bushnell offered Jobs a bonus and Atari would pay him a base fee plus $100 for every chip eliminated from the design (fewer chips meant a simpler, cheaper circuit). Jobs told Wozniak that Atari would pay them a flat amount, and that they would split it 50/50. Wozniak, eager to help and excited by the challenge, worked tirelessly for four nights straight and came up with an incredibly efficient design that reduced the chip count dramatically. Atari was impressed and gave Jobs a large bonus reportedly around $5,000 (some accounts say $5,000, others $6,000). But Jobs never told Wozniak about the bonus. Instead, he gave Wozniak $350, telling him that was half of what Atari paid. Wozniak didn’t find out the truth until ten years later, when he read it in a book about Atari (Zap!). He was deeply hurt but chose to shrug it off.

It was the British Intelligence which tipped off Canada on alleged signit on Nijjar Hit setting off a diplomatic crisis between New Delhi & Ottawa even involving Washington ! Interesting

Hilarious that China is building these reactors and not USA. I wasted a lot of my youth campaigning online in English for fast reactors. I should have learnt Chinese instead. 😆

It's too easy to gas yourself up thinking about your percentile. "Woah, I'm 99th percentile, I'm such a baller..." like dude you're 1 in 100. In a country with 100 million people there's a million of you. You wanna get to be one in a million, you need to improve by many orders of magnitude. It's just too easy to think "99th... 99.9th... 99.99th... whatever they're basically same thing, 99th is nearly there" -- if you wanna have an accurate intuitive sense of how good you are, don't think about the percentile, think about how many of you there are.

Math is learned by doing! To use Prof. Halmos' words.... Don't just read it, fight it! Do the exercises, try different approaches, get frustrated, write possible solutions on coffee napkins, take a break, go for a walk and ponder... Repeat. 🙏💪😎

What?

I'm neither Bengali nor liberal lol. But it is revealing to see how quickly other Indians resort to anti-Bengali bigotry when they disagree with me, just because I have the Bangla language in my X username 🤷‍♂️