HN Explorer

ashvardanian

↗ HN profile [ 66.7 ms ] full profile
Karma
0
Created
()
Submissions
0
  1. AI Model Co-Design: Hardware-Friendly LLM Design (developer.nvidia.com)
    1 points 0
  2. KinetIQ Ascend: Toward 100% Reliable Manipulation and Superhuman Speed (thehumanoid.ai)
    5 points 2
  3. Nvidia CUDA Python 1.0 and CUDA 13.3 Release (developer.nvidia.com)
    2 points 2
  4. Laguna XS.2 and Laguna M.1 by Poolside (poolside.ai)
    5 points 0
  5. FP8 Search and KV-Caching in USearch (unum.cloud)
    1 points 0
  6. Escaping the Fork: How Meta Modernized WebRTC Across 50 Use Cases (engineering.fb.com)
    3 points 0
  7. Porting Go's io package to C (antonz.org)
    6 points 0
  8. Schema as the Core of Reliability in AI Memory (xmemory.ai)
    6 points 0
  9. NumKong: 2'000 Mixed Precision Kernels for All (ashvardanian.com)
    47 points 3
  10. The State of Allocators in 2026 (cetra3.github.io)
    2 points 0
  11. Recraft V4 (recraft.ai)
    2 points 0
  12. Nebius to buy AI agent search company Tavily for 275M (nebius.com)
    2 points 1
  13. Running Async WebAssembly on Seastar's Reactor (rockwotj.com)
    3 points 0
  14. Open source USearch library jumpstarts ScyllaDB vector search (thenewstack.io)
    2 points 0
  15. Mixedbread: How We Built Multimodal Late-Interaction at Billion Scale (mixedbread.com)
    3 points 0
  16. Nvidia: Using Context as Training Data Unlocks Models That Learn at Test-Time (developer.nvidia.com)
    6 points 0
  17. Nvidia Kicks Off the Next Generation of AI with Rubin – Six New Chips (nvidianews.nvidia.com)
    15 points 2
  18. Building Code-Chunk: AST Aware Code Chunking (supermemory.ai)
    2 points 0
  19. Rust: Using String and andstr in your APIs (dx13.co.uk)
    3 points 1
  20. 1 points 0
  21. From Ts_rank to BM25. Introducing Pg_textsearch: True BM25 Ranking and Retrieval (tigerdata.com)
    2 points 0
  22. Full Unicode Search at 50× ICU Speed with AVX‑512 (ashvardanian.com)
    217 points 18
  23. Emulating AVX-512 intrinsics in Miri (trifectatech.org)
    14 points 0
  24. Why we built Lightpanda in Zig (lightpanda.io)
    212 points 27
  25. Nvidia cuTile: Python DSL and a new IR for tile-based CUDA kernels (github.com)
    5 points 0