← 최신 프로덕트

UL-SMF

Open-source linear-complexity ~300x KV-cache compression. The Unified Latent-State Memory Fabric (UL-SMF) is a hardware-software co-designed memory compression fabric that solves the memory bottleneck in long-context Transformer inference. By combining FSQ with dynamic 16-dimensional latent mapping, UL-SMF compresses Key-Value (KV) cache tensors by up to 384x while maintaining >94% semantic retention.

오픈소스

10 스타 · GitHub외부 사이트

github.com 방문외부 사이트

반응 (1)

비슷한 프로덕트

설명이 비슷한 순

  1. UltraCompress

    github.com

    first mathematically lossless 5-bit LLM compression

    Hacker News6점

  2. Micro-RLE ≤264-byte compression for UART/MCU logs

    github.com

    zero RAM growth

    Hacker News7점

  3. Fixing LLM memory degradation in long coding sessions

    github.com

    Empirical proof of SOTA LLM (GPT-5/Gemini-Pro/Claude-Pro) context saturation in complex engineering. Contains the "Misuraca Protocol" for deterministic logical segmentation to prevent entropy drift.

    Hacker News5점

  4. Cuts Long Horizon Inference Costs by 50% via external KV Cache Offload

    github.com

    OpenLake is a high performance storage engine for efficient LLM inference and GPU Training

    Hacker News22점

  5. Kanzi

    github.com

    fast lossless data compression

    Hacker News15점