Qwen3.8-Flash-Next (W4A16) on 2x CMP 170HX: 110 tok/s decode, 7K prefill, 800 tok/s @ 23 concurrent - runs on 38 GB home RAM, no pinned memory
☆20Sep 18, 2026Updated last month
Alternatives and similar repositories for qwen38-flash-next-w4a16-cmp170hx
Users that are interested in qwen38-flash-next-w4a16-cmp170hx are comparing it to the libraries listed below. We may earn a commission when you buy through links labeled 'Ad' on this page.
Sorting:
- 双 SM80 / CMP 170HX 的 Qwen3.8-Flash-Next 社区 vLLM：NVIDIA NVFP4、GDS、TEP2、MTP6，支持 1～10 并发，词表与 PCIe 通信优化。中英文部署与可构建源码。☆28Oct 15, 2026Updated 3 weeks ago
- dsv41-flash-pp5-170hx &pp6 Supports the original (full/unquantized) model. Prefill 7,000 t/s, decode 45-75-125 t/s. dspark. Requires 200 …☆13Updated this week
- Linux P2P unlock and performance fixes for NVIDIA CMP 170HX☆39Sep 22, 2026Updated last month
- 限制问题区域显存 62906MiB / 65024MiB 支持驱动 610.57.04 ，修改显卡显示名称为NVIDIA Tesla A100☆12Aug 30, 2026Updated 2 months ago
- A cmpunlocker fork with working p2p on SYS-4028GR-TRT2☆27Aug 8, 2026Updated 3 months ago
- Deploy on Railway without the complexity - Free Credits Offer • AdConnect your repo and Railway handles the rest with instant previews. Quickly provision container image services, databases, and storage volumes.
- DeepSeek-V4.1-Flash (764B) on 4×CMP 170HX: PP4 + shadow KV + EXL3 2bpw + DSpark + 512k ctx + native vision — patches, recipes, acceptance…☆15Sep 29, 2026Updated last month
- A tool to unlobotomize your NVIDIA card!☆841Oct 28, 2026Updated 2 weeks ago
- some experiments about 170hx 10g version.☆11Oct 19, 2026Updated 3 weeks ago
- vLLM 0.33 fork for consumer GPUs: 3 x 262K sessions on 4x RTX 3090, Qwen 3.8 Flash-Next at 207 tok/s, up to 17 on bigger cards. Seven che…☆49Updated this week
- Unlock tool for the NVIDIA CMP 170HX (GA100) mining card. Restores full A100 compute throughput by exploiting the Falcon BootROM .fwsigna…☆57Jul 30, 2026Updated 3 months ago
- 适配泰安S8030显卡机自动调节系统风扇组件☆13Dec 13, 2024Updated last year
- Exact LLM inference engine written from scratch for AMD Strix Halo (gfx1151, ROCm/HIP). Qwen3.8-Flash-Next on one node or two replicas; D…☆23Updated this week
- Auto-built Docker image of 1Cat-vLLM for NVIDIA Tesla V100 (sm_70/Volta) with CUDA 12.8 — rebuilt daily when upstream releases☆12Jul 22, 2026Updated 3 months ago
- 在四卡A100上适配Sglang部署deepseek-v4-flash正式版模型☆15Aug 2, 2026Updated 3 months ago
- Simple, predictable pricing with DigitalOcean hosting • AdAlways know what you'll pay with monthly caps and flat pricing. Enterprise-grade infrastructure trusted by 600k+ customers.
- DeepSeek V4.1 Flash 1017 MXFP4 on one RTX PRO 5000 72GB: 512K shared KV, 2 concurrent requests, DSpark and compiled lkqmoe 0.5.4.☆16Oct 24, 2026Updated 2 weeks ago
- Active4j-jsp是基于SpingBoot2.0轻量级的java快速开发框架。以Spring Framework为核心容器，Spring MVC为模型视图控制器，Mybatis Plus为数据访问层， Apache Shiro为权限授权层, Redis为分布式缓存，Q…☆15Feb 11, 2026Updated 9 months ago
- The original as far as I could find.☆14Feb 10, 2025Updated last year
- Fast vLLM on AMD Strix Halo: one docker command for Qwen, Gemma, LFM and more, up to 31x faster prefill and ~2x decode vs stock vLLM at l…☆21Updated this week
- vLLM for NVIDIA CMP 170HX (Ampere sm_80): fast, repeatable model serving on one card or several☆28Nov 2, 2026Updated last week
- A deterministic inference engine for DeepSeek-V4.1-Flash 1017 and Qwen3.8-Flash-Next on a single AMD Strix Halo 128 GB☆22Updated this week
- forked from vllm-project/flash-attention☆74May 9, 2026Updated 6 months ago
- ☆25Jun 27, 2026Updated 4 months ago
- A tool to unlobotomize your NVIDIA card!☆31Sep 6, 2026Updated 2 months ago
- Managed hosting for WordPress and PHP on Cloudways • AdManaged hosting for WordPress, Magento, Laravel, or PHP apps, on multiple cloud providers. Deploy in minutes on Cloudways by DigitalOcean.
- Container toolbox that runs llama.cpp with AMD ROCm and NVIDIA CUDA in one build, for inference on two GPUs☆22Oct 16, 2026Updated 3 weeks ago
- Dragonflight_Dekk-Core☆20Dec 23, 2023Updated 2 years ago
- Docker Compose for llama.cpp GGUF servers on AMD Strix Halo: Qwen, Gemma, and Laguna packages (abliterated and quantized), stock Vulkan p…☆28Aug 17, 2026Updated 2 months ago
- Pong in Haskell / Clash, running as software using SDL and as hardware targeting FPGAs☆10Apr 20, 2024Updated 2 years ago
- MCP server for file operations on non-UTF-8 text: detects the encoding from the bytes, hands the model UTF-8, and writes back in the orig…☆30Updated this week
- ☆21Jan 31, 2025Updated last year
- Qwen3.8-27B via vLLM on a single AMD Radeon AI PRO R9700 (gfx1201, 32 GiB) — bare, script-free setup guide, written to be executed by an …☆13Sep 3, 2026Updated 2 months ago
- OpenCore EFI for X79 Mainboard and CPU xeon E5 2650v2 RX470☆20Feb 1, 2021Updated 5 years ago
- The core18 base snap☆14Updated this week
- Wordpress hosting with auto-scaling - Free Trial Offer • AdFully Managed hosting for WordPress and WooCommerce businesses that need reliable, auto-scalable performance. Cloudways SafeUpdates now available.
- vLLM on AMD RDNA4 (gfx1201 / RX 9070 XT-9070): clone-and-docker-compose-up, with a W4A8-FP8-WMMA MoE kernel☆14Jul 18, 2026Updated 3 months ago
- Dogfooding vLLM backport for older GPUs.☆362Updated this week
- ☆44Nov 2, 2026Updated last week
- A series of guides on how to setup devices that are Arm SystemReady certified and their firmware.☆15Feb 25, 2022Updated 4 years ago
- ☆11Sep 15, 2017Updated 9 years ago
- Doppler effect on WaveForms☆17Sep 1, 2025Updated last year
- Container driven LLM control for VLLM & llama models☆14Updated this week
Are these results useful?