Skip to content
View aktalascloud's full-sized avatar

Block or report aktalascloud

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
aktalascloud/README.md

Graphite processor with a copper lightning bolt

FeraVolt · Local AI on older hardware

Русский · Hugging Face · TalasCloud

I optimize local LLM inference and generative media for low-spec PCs. My projects include portable Windows builds and GGUF quantization. I work with NVIDIA Pascal and Tesla P40 over CUDA, AMD Polaris and RX 580 over Vulkan, Intel Ivy Bridge with AVX1, and Haswell with AVX2.

Inference and model tools

Project Use it for
Strata Desktop LLMs on Tesla P40 · local Swift 256K, CUDA and AVX1
pascallama.cpp llama.cpp DP4A / MMQ patches for Pascal
KoboldCpp Core Haswell AVX2 + Polaris Vulkan runtime
Polaris AI Toolkit RX 580 toolkit for ZLUDA, HIP/ROCm and ComfyUI
GGUF Converter Portable Windows conversion + Vulkan calibration
KoboldCpp Media Local API for images, video, music and sound

Interfaces and agents

Project Use it for
Open WebUI AVX1 Self-hosted interface for older Windows PCs
Open WebUI Desktop AVX1 Desktop client with isolated connections
Open Terminal AVX1 Standalone terminal and agent service
Conduit ARM64 Open WebUI + native Cline on Android
Cline Desktop AVX1 Archived Ivy Bridge Windows client

Measured results

  • Tesla P40 · Strata. One Swift IQ3_S request with 261,998 input tokens and 128 output tokens completed in 913.43 s without a cache read. Evidence
  • RX 580 · Polaris. Synthetic 4096×4096 attention took 87.68 → 53.86 ms and 1,024 → 69 MiB in additional peak tensor allocations. Evidence
  • RX 580 / Haswell · Media. LightVAE-Pro saved 17 Wan frames in 30.53 s vs 33.04 s. The SA3 CPU worker processed three chunks of a 15-second effect in 21.57 s vs 26.53 s. Outputs matched byte for byte. Evidence

Models on Hugging Face

Calibrated GGUF, ONNX exports and original components. Model cards include upstream revisions, licenses, hashes and compatible runtimes.

Images / Изображения · Video / Видео · Music / Музыка · Sound effects / Звуковые эффекты

Popular repositories Loading

  1. polaris-gfx803-rx580-zluda-rocm-hip-comfyui-windows polaris-gfx803-rx580-zluda-rocm-hip-comfyui-windows Public

    AMD Polaris RX 580 / gfx803 Windows AI toolkit: ZLUDA, HIP/ROCm, rocBLAS, DirectML, optional ComfyUI. / AI-стек для RX 580 на Windows.

    Python 4

  2. koboldcpp-avx2-vulkan-numa koboldcpp-avx2-vulkan-numa Public

    Forked from LostRuins/koboldcpp

    KoboldCpp for Intel Haswell AVX2 and AMD RX 580 Polaris Vulkan. / Портативный KoboldCpp для Haswell, RX 580 и локальных GGUF-моделей.

    C++ 2

  3. koboldcpp-media-pipeline koboldcpp-media-pipeline Public

    Local Windows AI media API for MOSS, Wan, YuE2 and Stable Audio on Polaris/Haswell. / Локальный API для изображений, видео, музыки и звука.

    Python 2

  4. strata-legacy-gpu strata-legacy-gpu Public

    Forked from Niko1221/Strata

    Local LLM desktop for Tesla P40 Pascal, CUDA MMQ and Ivy Bridge AVX1. / Локальные LLM на Tesla P40 и AVX1, Swift 256K и LAN-медиа.

    C++ 2

  5. open-webui-ivybridge-avx1 open-webui-ivybridge-avx1 Public

    Forked from open-webui/open-webui

    Self-hosted Open WebUI for Windows, Intel Ivy Bridge AVX1, local LLM and MCP. / Open WebUI для Ivy Bridge, русский интерфейс и MCP.

    Svelte 2

  6. open-terminal-ivybridge-avx1 open-terminal-ivybridge-avx1 Public

    Forked from open-webui/open-terminal

    Open Terminal for Windows and Ivy Bridge AVX1, with isolated agent tools and bounded output. / Терминал для агентов на Ivy Bridge.

    Python 2 1