--- title: "My Open Software" date: 0001-01-01 source: https://ashvardanian.com/software/ author: Ash Vardanian --- All of my software is hosted on GitHub, mostly under the [Apache-2.0](https://opensource.org/licenses/Apache-2.0) permissive license. Free for commercial and non-commercial use, modification, and distribution. ## Major Projects - __[USearch](https://github.com/unum-cloud/USearch)__ - one of the world's most popular search engines, used in databases, AI labs, and large-scale Natural Science experiments. Compact C++ core with 10+ language bindings — 10–100× faster than Meta FAISS for vector search and far beyond Apache Lucene. - __[StringZilla](https://github.com/ashvardanian/StringZilla)__ - one of the fastest string/text processing libraries, leveraging SIMD, SWAR, and CUDA-accelerated algorithms for search, matching, hashing, and sorting at Web-scale Unicode UTF-8 text and Petabyte-scale Bioinformatics data. Hundreds of hand-tuned kernels with manual multi-versioning, exposed to C, C++, Rust, Python, Swift, and JavaScript, up to 10× faster on CPUs and 100× faster on GPUs. [CLI](https://github.com/ashvardanian/StringZilla-CLI) included. - __[NumKong](https://github.com/ashvardanian/NumKong)__ - one of the largest & most-portable mixed-precision numerics projects for C, C++23, Python, Rust, and JavaScript. Designed for linear algebra, scientific computing, statistics, information retrieval, and image processing, delivering consistent SIMD speedups over BLAS and NumPy on x86, ARM, RISC-V, PowerPC, LoongArch, and in-browser WASM environments, bringing 6-, 8-, and 16-bit obscure GPU-only floats to every modern chip. - __[ForkUnion](https://github.com/ashvardanian/ForkUnion)__ - ultra-low-latency parallelism library for Rust and C++. Avoids allocations, mutexes, and even Compare-And-Swap atomics — achieving up to 10× speedups over Rayon and TaskFlow. - __[UCall](https://github.com/unum-cloud/UCall)__ - a kernel-bypass web server backend for C and Python built on io_uring. Achieves 70× higher throughput and 50× lower latency than FastAPI for real-time workloads, including serving compact AI models. - __[UForm](https://github.com/unum-cloud/UForm)__ - tiny multimodal AI models with state-of-the-art parameter and data efficiency. Compatible with Python, JS, and Swift, serving as a lightweight alternative to OpenAI CLIP for on-device and server inference. Some of those are used in open-source databases, like [ClickHouse](https://github.com/ClickHouse/ClickHouse), [DuckDB](https://github.com/duckdb/duckdb), [TiDB](https://github.com/pingcap/tidb), [ScyllaDB](https://github.com/scylladb/scylladb), [yugabyteDB](https://github.com/yugabyte/yugabyte-db), [DragonflyDB](https://github.com/dragonflydb/dragonfly), [MemGraph](https://github.com/memgraph), [Vald](https://github.com/vdaas/vald), [Turso](https://github.com/tursodatabase/turso), [Chroma](https://github.com/chroma-core/chroma), LLM toolchains, like [LangChain](https://github.com/langchain-ai/langchain), [LlamaIndex](https://github.com/run-llama/semtools), [Microsoft SemanticKernel](https://github.com/microsoft/semantic-kernel), [Nomic AI GPT4All](https://github.com/nomic-ai/gpt4all), [Surf](https://github.com/deta/surf), and many other less "open" systems, such as backend infrastructure of major AI labs, government intelligence agencies, hyperscale cloud companies, Fortune 500, iOS and Android apps with 100M-1B MAU. ## Tutorials & Datasets - __[less_slow.cpp](https://github.com/ashvardanian/less_slow.cpp)__ - teaches a performance-oriented mindset for C++, CUDA, PTX, and ASM - __[less_slow.rs](https://github.com/ashvardanian/less_slow.rs)__ - Rust adaptation with a focus on higher-level abstractions - __[less_slow.py](https://github.com/ashvardanian/less_slow.py)__ - Python adaptation with a focus on scripting & data-management - __[SpaceV](https://github.com/ashvardanian/SpaceV)__ - 1 billion vectors from Microsoft SpaceV extended for usability - __[USearchMolecules](https://github.com/unum-science/USearchMolecules)__ - 28 billion fingerprints for drug discovery, published with AWS ## Demos & Benchmarks - [RetriEval](https://github.com/ashvardanian/RetriEval) - billion-scale Vector Search benchmarks for USearch, FAISS, cuVS, Weaviate, Qdrant, etc. - [SmashTable](https://github.com/ashvardanian/SmashTable) - in-memory collections with DBMS-like consistent & atomic transactions in Python and C++ - [NumWars](https://github.com/ashvardanian/NumWars) - micro-benchmarking NumKong against the best Rust & Python BLAS libraries - [StringWars](https://github.com/ashvardanian/StringWars) - micro-benchmarking StringZilla against the best Rust & Python string libraries - [HashEvals](https://github.com/ashvardanian/HashEvals) - testing avalanche effect & differential patterns of string hash functions - [ScalingElections](https://github.com/ashvardanian/ScalingElections) - parallel combinatorial voting in CUDA and Mojo for H100 GPUs - [TinySemVer](https://github.com/ashvardanian/TinySemVer) - semantic versioning GitHub CI tool that doesn't take 300K lines of JavaScript - [SwiftSemanticSearch](https://github.com/ashvardanian/SwiftSemanticSearch) - example of on-device real-time AI using UForm and USearch on iOS - [ParallelReductionsBenchmark](https://github.com/ashvardanian/ParallelReductionsBenchmark) - GPGPU benchmarks for SyCL, CUDA, OpenCL, Vulkan, etc. - [LibSee](https://github.com/ashvardanian/libsee) - non-intrusively profiling LibC calls with `LD_PRELOAD` tricks - [StringTape](https://github.com/ashvardanian/StringTape) - Apache Arrow compatible tapes for space-efficient string arrays - [JsonTape](https://github.com/ashvardanian/JsonTape) - allocator-aware & comments-preserving JSON & JSON5 library for Rust - [JaccardIndex](https://github.com/ashvardanian/JaccardIndex) - exploring CPU port utilization with Carry-Save Adders & Lookups - [UStore](https://github.com/unum-cloud/UStore) - multimodal embedded database for C, C++, and Python designed around key-value stores - [SmallField](https://github.com/unum-science/SmallField) - small-molecule 3D conformers generation - on GPUs & in mixed-precision - [AffineGaps](https://github.com/unum-science/AffineGaps) - "less wrong" local and global Gotoh sequence alignments in one NumBa Python file - [FasterFASTA](https://github.com/unum-science/FasterFASTA) - CLI tool to parse, sort, dedup, and translate DNA, RNA, & protein sequences