# building block (blogs) — RSS Amplifier Recent posts from the 12 feeds in the RSS Amplifier directory that cover building block. Page: Feed: --- ## [Anaheim hotel renovation brings new suites, outdoor upgrades](https://www.constructionspecifier.com/anaheim-hotel-renovation-brings-new-suites-outdoor-upgrades/) _2026-08-18 · amrit\_bls · Construction Specifier_ R.D. Olson Construction has started renovations at the Howard Johnson Anaheim Hotel and Water Playground. Explore the project. The post Anaheim hotel renovation brings new suites, outdoor upgrades appeared first on Construction Specifier . ## [Renovated facility expands access to specialized care](https://www.constructionspecifier.com/renovated-facility-expands-access-to-specialized-care/) _2026-08-18 · amrit\_bls · Construction Specifier_ Find out how Skanska used coordinated construction methods to maintain production and control resources despite permitting challenges. The post Renovated facility expands access to specialized care appeared first on Construction Specifier . ## [Play Building Blocks: Wordplay for Specifiers—August 18, 2026](https://www.constructionspecifier.com/play-building-blocks-wordplay-for-specifiers-august-18-2026/) _2026-08-18 · amrit\_bls · Construction Specifier_ Try your hand at this week's edition of Building Blocks. Test your construction knowledge with The Construction Specifier’s new word game. The post Play Building Blocks: Wordplay for Specifiers—August 18, 2026 appeared first on Construction Specifier . ## [Introducing ALPOLIC™/NC-US, the U.S. Market’s First Non-combustible Metal Composite Material](https://www.constructionspecifier.com/introducing-alpolic-nc-us-the-u-s-markets-first-non-combustible-metal-composite-material/) _2026-08-18 · amrit\_bls · Construction Specifier_ Meet the next generation of beautiful, durable, and safe metal composite materials. Elevate your exterior with the new ALPOLIC™/NC-US. The post Introducing ALPOLIC™/NC-US, the U.S. Market’s First Non-combustible Metal Composite Material appeared first on Construction Specifier . ## [Trim-Tex Sound XL: The Better Caulk Alternative for Sound-Rated Walls](https://www.constructionspecifier.com/trim-tex-sound-xl-the-better-caulk-alternative-for-sound-rated-walls/) _2026-08-18 · amrit\_bls · Construction Specifier_ See why contractors are choosing Sound XL for high-performance sound and smoke ratings on head-of-wall joints. The post Trim-Tex Sound XL: The Better Caulk Alternative for Sound-Rated Walls appeared first on Construction Specifier . ## [Innovations of Solid 2.0](https://brenelz.com/posts/innovations-of-solid-2/) _2026-08-13 · Brenelz :: TypeScript/React Developer_ ## [Play Building Blocks: Wordplay for Specifiers—August 11, 2026](https://www.constructionspecifier.com/play-building-blocks-wordplay-for-specifiers-august-11-2026/) _2026-08-11 · amrit\_bls · Construction Specifier_ Put your construction knowledge to the test with the first edition of Building Blocks, The Construction Specifier’s new word game. The post Play Building Blocks: Wordplay for Specifiers—August 11, 2026 appeared first on Construction Specifier . ## [How to Play Building Blocks](https://www.constructionspecifier.com/how-to-play-building-blocks/) _2026-08-11 · amrit\_bls · Construction Specifier_ Introducing "Building Blocks," The Construction Specifier ’s new weekly word puzzle for AEC professionals. The post How to Play Building Blocks appeared first on Construction Specifier . ## [Design firm takes home significant Gold Nugget prize haul](https://www.constructionspecifier.com/design-firm-takes-home-significant-gold-nugget-prize-haul/) _2026-08-11 · amrit\_bls · Construction Specifier_ Architecture firm AO earned four Grand Awards and 24 Merit Awards at the 63rd annual Gold Nugget Awards. See which project were recognized. The post Design firm takes home significant Gold Nugget prize haul appeared first on Construction Specifier . ## [Scholarships worth $225,200 handed out](https://www.constructionspecifier.com/scholarships-worth-225200-handed-out/) _2026-08-11 · amrit\_bls · Construction Specifier_ The scholarships recognize students who show academic excellence and a commitment to careers in engineering and the built environment. The post Scholarships worth $225,200 handed out appeared first on Construction Specifier . ## [Generate Viral Hot Takes (Sponsored)](https://crawlproof.com/a/SpwUJ7gk6W8v) _2026-08-11 · **Sponsored**_ Get platform-optimized hot takes and rebuttals to spark debate and engagement ## [The Journey from a Vite SPA to Server Components](https://brenelz.com/posts/the-journey-from-a-vite-spa-to-server-components/) _2026-08-11 · Brenelz :: TypeScript/React Developer_ ## [GET TO SPEC FASTER: The New Steel Joist and Deck Experience](https://www.constructionspecifier.com/get-to-spec-faster-the-new-steel-joist-and-deck-experience/) _2026-08-10 · amrit\_bls · Construction Specifier_ After New Millennium completely rebuilt its digital experience, finding the right steel joist and deck information has never been easier. The post GET TO SPEC FASTER: The New Steel Joist and Deck Experience appeared first on Construction Specifier . ## [How I Do Software Development in July 2026](https://jasuja.us/2026/07/how-i-develop-software-in-july-2026/) _2026-07-23 · Nick Jasuja · (untitled)_ Six months ago in Jan I ported JustHTML to PHP. I had also released a CLI for it and published it via Homebrew to make document parsing, selecting DOM nodes and dumping to markdown super easy. Most of the heavy lifting at the time was done by ChatGPT 5.2 Codex, a fantastic model that did \[ \] ## [A Practical Guide to Reducing Token Spend](https://www.adamhjk.com/blog/a-practical-guide-to-reducing-token-spend/) _2026-07-16 · Adam Jacob_ Building a Swamp Workflow reduced real world token usage of an AI Agent by 8x and improved execution time by 2x - here is how to do it yourself. ## [Writing A Megakernel For LLM Decode - A Worklog](https://emre570.bearblog.dev/megakernel-decode/) _2026-07-02 · emre570 · Emre's Blog_ This is the story of a megakernel: the 28-layer text decoder of Qwen3-ASR-0.6B, plus its lm\_head, running as a single persistent CUDA kernel on an RTX 5080. The baseline fires about 466 kernel launches per generated token. This fires one. Every op in it is hand-written - the GEMVs, the norms, the rope, the KV-cache write, the attention, the sampling argmax - and every op was validated against… ## [How My Mental Model of \`isPending\` Changed](https://brenelz.com/posts/is-pending-mental-model/) _2026-06-10 · Brenelz :: TypeScript/React Developer_ ## [Handling Errors in Solid 2.0](https://brenelz.com/posts/handling-errors-in-solid-2/) _2026-05-25 · Brenelz :: TypeScript/React Developer_ ## [Learning Solid 2.0](https://brenelz.com/posts/learning-solid2/) _2026-05-21 · Brenelz :: TypeScript/React Developer_ ## [Unleashing Blackwell's 4-bit: a surgical look at MXFP4 and NVFP4](https://emre570.bearblog.dev/blackwell-fp4/) _2026-05-19 · emre570 · Emre's Blog_ If you do kernel-level inference optimization, you eventually hit the cold truth: the enemy is not compute, it is VRAM bandwidth. FP32 numbers are huge in memory. The fix is to squeeze them into 4-bit boxes. The internet is full of repos that do this. They say "find amax, divide by this, shift the bits, here is your FP4," and move on. I copied those formulas into my own code at first - and could… ## [From 429 GB/s to the DRAM wall: writing an FP8 quantizer on an RTX 5080](https://emre570.bearblog.dev/fp8-quantizer/) _2026-05-13 · emre570 · Emre's Blog_ This is the story of one CUDA kernel pair in cublade, my personal kernel library. The kernels do per-tensor symmetric FP8 (E4M3) quantization and dequantization. They started at 429 GB/s on a v1 build, then NCU shoved them to roughly 880 GB/s at the kernel level - about 98.5% of the RTX 5080's real DRAM peak. Plus a dequantizer that landed DRAM-bound on its first compile. I'm going to walk through… ## [Home Internet From Space (Sponsored)](https://crawlproof.com/a/8zSrmKmjYHX4) _2026-05-13 · **Sponsored**_ Residential satellite internet that brings connectivity to homes beyond conventional networks ## [8.5x Faster Speech-to-Text: From 429ms to 50ms on a Single GPU](https://emre570.bearblog.dev/stt-blog-8x/) _2026-05-07 · emre570 · Emre's Blog_ My STT journey started with Whisper. One of Freya's STT models was a fine-tuned Whisper Large-V3. I optimized it with TensorRT, got it to 130ms, around 94x realtime. Solid, but we hit a wall. The architecture has a heavy encoder and a lightweight decoder, and TRT could only help with the encoder side. We needed something better. STT is the first link in our voice agent chain: STT → LLM → TTS. When… ## [Laziness, Impatience, and Hubris](https://www.adamhjk.com/blog/laziness-impatience-hubris/) _2026-04-14 · Adam Jacob_ How Larry Wall's virtues of a Perl programmer inspire me in this era of engineering ## [Adaptive Building Blocks](https://www.adamhjk.com/blog/adaptive-building-blocks/) _2026-04-09 · Adam Jacob_ Adaptive building blocks are how automation keeps up with AI development ## [Communication Breakdown](https://www.adamhjk.com/blog/communication-breakdown/) _2026-04-07 · Adam Jacob_ Talking with Stephen O'Grady about AI and building the machine that builds the machine ## [Avoiding the Lethal Trifecta for Personal Agents](https://www.adamhjk.com/blog/avoiding-the-lethal-trifecta-with-swamp/) _2026-03-31 · Adam Jacob_ How teaching your Claw about Swamp can help avoid the lethal trifecta ## [You still have to refactor, even with AI](https://www.adamhjk.com/blog/you-still-have-to-refactor-even-with-ai/) _2026-03-29 · Adam Jacob_ Engineering with AI Agents doesn't mean you get to skip the fundamentals ## [Things Learned Migrating To Solid 2.0](https://brenelz.com/posts/migrating-to-solid-2/) _2026-03-29 · Brenelz :: TypeScript/React Developer_ ## [As we build, so we believe](https://www.adamhjk.com/blog/as-we-build-so-we-believe/) _2026-03-28 · Adam Jacob_ Building the machine that builds the machine, in order to learn what I believe about the future of my profession ## [The US-Israeli attack on Iran](https://jasuja.us/2026/03/the-us-israeli-attack-on-iran/) _2026-03-24 · Nick Jasuja · (untitled)_ There was no reason for the US to attack Iran. What boggles the mind is that the administration is not even bothering to justify the attack. The Bush/Cheney administration at least had the decency to lie to the public about weapons of mass destruction so they could manufacture a pretext for the Iraq war. The \[ \] ## [Python Dictionaries and Tuples](https://kevin-haas.com/posts/python-dicts-and-tuples/) _2026-03-13 · xransum · Kevin Haas - Blog_ Overview You already know lists. They're great for ordered collections of things – a shopping cart, a sequence of numbers, a set of results. But they're not the right tool for every job. This post covers two more data structures. First dictionaries – the most important data structure in Python, and the one you'll use constantly. Then tuples – which you'll actually encounter inside dict... ## [Open directory of blogs (Sponsored)](https://crawlproof.com/a/qT3eBzdBpqLU) _2026-03-13 · **Sponsored**_ Drop URLs or OPML — get per-blog pages plus machine-readable JSON, OPML, text ## [Python Challenges 01](https://kevin-haas.com/posts/python-challenges-01/) _2026-03-13 · xransum · Kevin Haas - Blog_ Overview You've got the building blocks. Now put them to use. Each challenge below gives you a problem, the expected output, and a starter editor to work in. There's no single right answer – if your code produces the correct output, it's correct. Once you're done (or if you're stuck), expand the solution to see two different approaches and how they compare. Challenge 1: FizzBuzz Lo... ## [Python Basics](https://kevin-haas.com/posts/python-basics/) _2026-03-07 · xransum · Kevin Haas - Blog_ Overview Now that you've got Python installed, it's time to actually learn the language. This post covers the fundamentals you'll use in pretty much every Python program you ever write. Nothing too crazy yet – just the building blocks. Variables & Data Types Variables are how your program remembers things. You give a value a name, and from that point on you can use that name anywh... ## [What I learned porting JustHTML to PHP with GPT 5.2 Codex](https://jasuja.us/2026/01/porting-justhtml-to-php-with-gpt-5-2-codex/) _2026-01-16 · Nick Jasuja · (untitled)_ A couple of weeks ago, I read Emil Stenström’s write-up on building JustHTML, then Simon Willison’s post about porting it to JavaScript with Codex. That inspired me to create a PHP port in the same vein: a pure PHP library for HTML parsing that passes the html5lib test suite and stays compatible with as many PHP versions \[ \] ## [Solid.js Best Practices](https://brenelz.com/posts/solid-js-best-practices/) _2026-01-12 · Brenelz :: TypeScript/React Developer_ ## [Python REPL (in-browser)](https://kevin-haas.com/posts/python-repl-in-browser/) _2025-12-23 · xransum · Kevin Haas - Blog_ Python REPL (in-browser) This is a simplified integration of a Python REPL (Read-Eval-Print Loop) using PyScript. You can type Python code into the terminal below and it will run directly in your browser – no installation needed. ## [Introduction to Python](https://kevin-haas.com/posts/introduction-to-python/) _2025-12-23 · xransum · Kevin Haas - Blog_ What's the Purpose of This? A lot of people I talk to want to get into programming but don't know where to start. Python is always my first recommendation — it's simple enough to pick up quickly, but powerful enough to take you a long way. Why Python? Python is an accessible language due to its readable syntax and reduced emphasis on low-level concerns when compared to languages such as C, C... ## [Flash Attention from Scratch: Appendix B - Block Size Configuration](https://lubits.ch/flash/Appendix-B---Block-Size-Configuration) _2025-11-08 · Sonny's Blog_ This appendix dives into how block size configurations affect instruction patterns and performance in Flash Attention. ## [Flash Attention from Scratch Part 8: Instruction Reduction](https://lubits.ch/flash/Part-8) _2025-11-08 · Sonny's Blog_ Intro In Part 6, we improved our kernel to slightly outperform the reference kernel on the RTX 3090, but found that on the A100, it only reached 80.3% of the reference. ## [Appendix A - Ampere Microarchitecture](https://lubits.ch/flash/Appendix-A---Ampere-Microarchitecture) _2025-11-01 · Sonny's Blog_ Ampere Microarchitecture This part dives into the SM architecture to reveal how execution units compete for resources. ## [Flash Attention from Scratch Part 7: A100 Profiling](https://lubits.ch/flash/Part-7) _2025-11-01 · Sonny's Blog_ In Part 6, we exceeded reference performance on the RTX 3090, hitting 101.5% through FP instruction fusion and auto-tuning. ## [Flash Attention from Scratch Part 6: FP Instruction Fusion and Auto-Tuning](https://lubits.ch/flash/Part-6) _2025-11-01 · Sonny's Blog_ In the previous part, we implemented three major optimizations from the CUTLASS GEMM library: eager block loading, sub-tiling with fragment interleaving, and double buffering. ## [W8A16 Quantization with LLM.int8-Style Outlier Handling](https://emre570.bearblog.dev/w8a16-quantization/) _2025-10-18 · emre570 · Emre's Blog_ In Phase 1 , we demonstrated that outliers in transformer models are not "statistical noise"—they are critical carriers of model intelligence and context. When we zeroed outliers in weight matrices, model outputs degraded catastrophically, producing repetitive loops and nonsensical text. But here's the problem : Quantizing weights from BF16 to INT8 reduces memory by ~50%, but it also destroys… ## [Flash Attention from Scratch Part 5: Cutlass GEMM Optimizations](https://lubits.ch/flash/Part-5) _2025-09-14 · Sonny's Blog_ Intro In the previous part, we implemented swizzling and achieved a dramatic 2x performance improvement by eliminating bank conflicts. ## [Flash Attention from Scratch Part 4: Bank Conflicts & Swizzling](https://lubits.ch/flash/Part-4) _2025-09-14 · Sonny's Blog_ Intro In the last part, we used the instructions covered in Part 2 to construct our first kernel and reached nearly half the performance of the official implementation on the RTX 3090. ## [How Critical Are Outliers in Transformer Models? A Live Experiment - Phase 1](https://emre570.bearblog.dev/outlier-experiment/) _2025-09-06 · emre570 · Emre's Blog_ Introduction Many assume that large language models (LLMs, Transformers) simply “memorize data” due to their scale. In reality, some part of their "intelligence" and creative capacity is packed into outlier values, extremely large or small numbers in their weight matrices and activation tensors. In this post, we’ll zero out these outliers and directly observe how this disrupts the model’s output… ## [My Chrome extension to follow/block users on Hacker News](https://jasuja.us/2025/08/my-chrome-extension-to-follow-block-users-on-hacker-news/) _2025-08-22 · Nick Jasuja · (untitled)_ Took some time off work to recharge this summer and built this: https://chromewebstore.google.com/detail/hn-followblock/dkbnpfgpgljemihbcdmadffanebbjacc It s my first Chrome extension. Largely built by Claude Code honestly. I just wanted to understand the process of building an extension, testing it locally in dev mode, and publishing it to the Chrome official store. Spoiler: it costs a one-time… ## [Flash Attention from Scratch Part 3: Kernel 1](https://lubits.ch/flash/Part-3) _2025-08-09 · Sonny's Blog_ Intro In Part 2, we explored the fundamental CUDA building blocks - tensor core operations (mma) and efficient memory transfers (cp.async & ldmatrix). ## [Flash Attention from Scratch Part 2: Building Blocks](https://lubits.ch/flash/Part-2) _2025-08-09 · Sonny's Blog_ Intro In this part, we ll explore the CUDA operations that form the foundation of our Flash Attention kernel. ## [Flash Attention from Scratch Part 1: Intro](https://lubits.ch/flash/Part-1) _2025-08-09 · Sonny's Blog_ Intro In this 10-part series, we re going to implement Flash Attention 2 from scratch on Ampere GPUs. ## [Mutable Reactivity (Diving into Stores)](https://brenelz.com/posts/mutable-reactivity/) _2025-06-20 · Brenelz :: TypeScript/React Developer_