README.md

Welcome to the HEC-Commander Blog!

Here are the latest blog posts with brief summaries:

Avoiding The Bitter Lesson in HEC-RAS Modeling

This article draws parallels between breakthroughs in AI and the computational challenges faced in hydraulic modeling using HEC-RAS, particularly in the 2D modeling era post version 6.0. Inspired by Rich Sutton's essay, *"The Bitter Lesson,"* the post explores how scaling through parallelism and brute force, rather than relying solely on smarter algorithms, can lead to significant improvements in modeling efficiency. It discusses the rising computational demands since HEC-RAS 5.0.7, the exhaustion of Moore’s Law, the challenges with parallelism in HEC-RAS, and how tools like RAS-Commander have addressed the gap by enabling up to 99 parallel runs. The post also provides insights on optimizing within current technological constraints, preparing for future GPU-based solvers, and embracing a new paradigm in hydraulic modeling.

Deep Dive: HEC-RAS 2D Infiltration Base Overrides using H5py

This blog post does a deep dive on the RAS-Commander code that modifies 2D infiltration base overrides in individual geometry files. Also includes the full code segments and examples of how to use the core functions in a user's own python workflows.

Benchmarking is All You Need: A Data-Driven Approach to Optimizing HEC-RAS Performance

This blog post details the informal benchmarking findings from 2022 supporting the previous blog posts about 10x engineering, cloud costs, and the bootlegger/hotrod mindset to HEC-RAS modeling hardware. Detailed benchmarking results, figures, and raw data files are included.

Courant Rules Everything Around Me: Balancing Accuracy, Resolution, and Efficiency in Large-Scale HEC-RAS Modeling

This blog post highlights the need for balancing cell size and time step in large scale HEC-RAS models, as well as common pitfalls of over-reliance on adaptive timestep. Anyone with a model that takes more than 24 hours should give this a read.

Using Terrain Mods on Your LIDAR Defined Channels

This blog post highlights a useful tip for achieving significant performance gains in HEC-RAS models by using RASMapper's terrain modifications layer to create pilot channels in LIDAR defined channels. It also introduces a new script, Terrain Mod Profiler, to automate the process of generating terrain profiles.

Thinking like a Bootlegger: Why Your HEC-RAS Modeling Machine Should be a Hot Rod, Not a Semi-Truck

This article explores the fundamentals of building a Hot Rod modeling machine. Whether you want to get the fastest HEC-RAS machine possible, or you have hundreds of runs to complete, the answer might not be what the salespersons try to sell you. Before making your next capital outlay for HEC-RAS compute, give this blog a read!

Moore's Law, AI Speedrunning the Golden Age of Computing

This article revisits the golden age of computing and explores the current resurgence in computational efficiency and power, driven by AI advancements and Moore's Law.

From 10x to 0.25x Engineering: By The Numbers

In this article, we delve into the quantitative aspects of 10x engineering. It provides a detailed analysis of how metrics and data-driven approaches are shaping the future of engineering, as well as a breakdown of cloud computing costs for various platforms and approaches.

10x Engineering in Water Resources with AI

This post discusses the role of AI in revolutionizing water resource management and engineering. It explores how AI can bring about a tenfold improvement in efficiency and effectiveness in this field.

Coming Soon:

Jupyter Notebooks: A LLM's Native Language for Code Execution

Automating Parallel HEC-RAS Execution: Command Line is All You Need

Automating HEC-HMS Execution with Jython: Jupyter Notebook Execution with Custom Parameters

Simplifying Python Notebooks Environment Setup with a Custom GPT

IDE-Based CoPilots: Novel Methods for Scripting with AI

Prompt Examples for Water Resource Engineers using HEC-RAS and HEC-HMS

Back to Basics: Optimizing 2D HEC-RAS Model Runtimes

Back to Basics: Optimizing 1D HEC-RAS Finite Volume Runtimes

Back to Basics: Optimizing Your Technology Platform for HEC-RAS Modeling