# Ash's Blog - [HackerNews Favorites](https://ashvardanian.com/hackernews/) - [My Open Software](https://ashvardanian.com/software/) - [Recordings & Talks](https://ashvardanian.com/talks/) ## Posts - [NumKong: 2'000 Mixed Precision Kernels For All 🦍](https://ashvardanian.com/posts/numkong/) - [Full Unicode Search at 50Γ— ICU Speed with AVX‑512](https://ashvardanian.com/posts/search-utf8/) - [Tuning TLS: AES-256 Beats ChaCha20 on Every CPU](https://ashvardanian.com/posts/chacha-vs-aes-2025/) - [Scaling Elections with GPUs and Mojo across Nvidia and AMD πŸ”₯](https://ashvardanian.com/posts/scaling-elections/) - [2x Faster Hashes on AWS Graviton: NEON β†’ SVE2](https://ashvardanian.com/posts/aws-graviton-checksums-on-neon-vs-sve/) - [Before AI's Kepler Moment - Are LLMs the Epicycles of Intelligence?](https://ashvardanian.com/posts/llm-epicycles/) - [How a String Library Beat OpenCV at Image Processing by 4x](https://ashvardanian.com/posts/image-processing-with-strings/) - [Processing Strings 109x Faster than Nvidia on H100](https://ashvardanian.com/posts/stringwars-on-gpus/) - [Beyond OpenMP in C++ & Rust: Taskflow, Rayon, Fork Union 🍴](https://ashvardanian.com/posts/beyond-openmp-in-cpp-rust/) - [CUDA Hello World: Done Less Wrong](https://ashvardanian.com/posts/less-wrong-cuda-hello-world/) - [The Longest Nvidia PTX Instruction](https://ashvardanian.com/posts/longest-ptx-instruction/) - [Hiding x86 Port Latency for 330 GB/s/core Reductions 🫣](https://ashvardanian.com/posts/cpu-ports/) - [Parsing JSON in C & C++: Singleton Tax](https://ashvardanian.com/posts/parsing-json-with-allocators-cpp/) - [10x Faster C++ String Split, 16 Years Later πŸ‘΄πŸ»](https://ashvardanian.com/posts/splitting-strings-cpp/) - [The Next 31 Years of Developing Unum](https://ashvardanian.com/posts/next-31-years-of-unum/) - [Understanding SIMD: Infinite Complexity of Trivial Problems πŸ”₯](https://ashvardanian.com/posts/understanding-simd-complexity/) - [5x Faster Set Intersections: SVE2, AVX-512, & NEON 🀐](https://ashvardanian.com/posts/simd-set-intersections-sve2-avx512/) - [35% Discount on Keyword Arguments in Python 🐍](https://ashvardanian.com/posts/discount-on-keyword-arguments-in-python/) - [NumPy vs BLAS: Losing 90% of Throughput](https://ashvardanian.com/posts/numpy-vs-blas-costs/) - [The Painful Pitfalls of C++ STL Strings 🧡](https://ashvardanian.com/posts/painful-strings/) - [USearch Molecules: 28 Billion Chemical Embeddings on AWS βš—οΈ](https://ashvardanian.com/posts/usearch-molecules/) - [Binding a C++ Library to 10 Programming Languages πŸ”Ÿ](https://ashvardanian.com/posts/porting-cpp-library-to-ten-languages/) - [Python, C, Assembly - 2'500x Faster Cosine Similarity πŸ“](https://ashvardanian.com/posts/python-c-assembly-comparison/) - [GCC Compiler vs Human - 119x Faster Assembly πŸ’»πŸ†šπŸ§‘β€πŸ’»](https://ashvardanian.com/posts/gcc-12-vs-avx512fp16/) - [Accelerating JavaScript arrays by 10x for Vector Search 🏹](https://ashvardanian.com/posts/javascript-ai-vector-search/) - [Our CPython bindings got 5x faster without PyBind11 🐍](https://ashvardanian.com/posts/pybind11-cpython-tutorial/) - [SciPy distances... up to 200x faster with AVX-512 & SVE πŸ“](https://ashvardanian.com/posts/simsimd-faster-scipy/) - [Combinatorial Stable Marriages for DBMS Semantic Joins πŸ’](https://ashvardanian.com/posts/searching-stable-marriages/) - [StringZilla: 5x faster strings with SIMD & SWAR πŸ¦–](https://ashvardanian.com/posts/stringzilla/) - [Abusing Vector Search for Texts, Maps, and Chess β™ŸοΈ](https://ashvardanian.com/posts/abusing-vector-search/) - [Counting Strings in C++: 30x Throughput Difference πŸ’¬](https://ashvardanian.com/posts/count-unique-strings/) - [We went through life with a smile πŸ’”](https://ashvardanian.com/posts/my-sona/) - [Mastering C++ with Google Benchmark ⏱️](https://ashvardanian.com/posts/google-benchmark/) - [Failing to Reach DDR4 Bandwidth 🚌](https://ashvardanian.com/posts/ddr4-bandwidth/) - [Crushing CPUs with 879 GB/s Reductions in CUDA](https://ashvardanian.com/posts/cuda-parallel-reductions/) - [Apple to Apple Comparison: M1 Max vs Intel 🍏](https://ashvardanian.com/posts/apple-m1/) - [Hyperscaler Shopping List: 2022 Data Center Tech Frenzy ☁️](https://ashvardanian.com/posts/server-supercycle/) - [Only 1% of Software Benefits from SIMD Instructions](https://ashvardanian.com/posts/simd-popularity/) - [Artsakh Must Be Independent πŸ—ΊοΈ](https://ashvardanian.com/posts/artsakh/) - [The 7 Sins of Turkish Autocracy πŸ‡ΉπŸ‡·](https://ashvardanian.com/posts/turkey/) - [Armenia, Azerbaijan, Turkey. Who's the Aggressor? βš”οΈ](https://ashvardanian.com/posts/aggressor/) - [Come to Armenia πŸ‡¦πŸ‡²](https://ashvardanian.com/posts/armenia/) - [Positive Outlook on the COVID-19 Crisis 😷](https://ashvardanian.com/posts/covid19/) - [Building AI Safely](https://ashvardanian.com/posts/building-ai-safely/) - [What's Wrong with WWDC 2016 Keynote?](https://ashvardanian.com/posts/whats-wrong-with-wwdc-2016/)