PulseAugur
EN
LIVE 13:45:34
العربية(AR) تشغيل نموذج MoE بحجم 35B على GTX 1650 بذاكرة 4GB: اختبار فعلي

35B MoE model runs on 4GB VRAM GPU via hybrid CPU/GPU inference

A technical test demonstrates that a 35-billion parameter Mixture of Experts (MoE) model, Qwen3.6-35B-A3B, can be run on a consumer-grade NVIDIA GTX 1650 Max-Q GPU with only 4GB of VRAM. The key to this feasibility is the MoE architecture, which uses only a subset of its parameters for each token, allowing the majority of the model's weights to be stored in system RAM rather than VRAM. This hybrid CPU/GPU inference approach achieved a practical speed of 21-22 tokens per second, making it usable for interactive applications. AI

IMPACT Demonstrates that large MoE models can be run on consumer hardware, potentially lowering the barrier to entry for advanced AI applications.

RANK_REASON Technical test demonstrating feasibility of running a large MoE model on limited hardware.

Read on dev.to — LLM tag →

AI-generated summary · Google Gemini · from 2 sources. How we write summaries →

35B MoE model runs on 4GB VRAM GPU via hybrid CPU/GPU inference

How we ranked this

Signal score
81 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Research
Technical test demonstrating feasibility of running a large MoE model on limited hardware.
Source corroboration
2 independent sources
Multiple independent publishers reporting the same story raises confidence that it's real and newsworthy.
Topics
infra, model release
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
Breaking (< 6h)
Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

Full methodology in our editorial standards.

COVERAGE [2]

  1. dev.to — LLM tag TIER_1 العربية(AR) · Nabil Doghri ·

    Running a 35B MoE Model on a 4GB GTX 1650: A Real-World Test

    <p>يبدو تشغيل نموذج يحتوي على 35 مليار مُعامل على بطاقة GTX 1650 Max-Q بذاكرة فيديو لا تتجاوز 4 غيغابايت مزحة سيئة. وغالبًا ما يكون كذلك.</p> <p>لكن النموذج الذي اختبرته من نوع «مزيج الخبراء» MoE، وهذا النوع لا يستخدم جميع معاملاته في حساب كل رمز. ومع ذاكرة نظام كافية، ومحرك استد…

  2. dev.to — LLM tag TIER_1 Français(FR) · Nabil Doghri ·

    Running a 35B MoE on a 4GB GTX 1650: The Real Test

    <p>Tester un modèle de 35 milliards de paramètres sur une GTX 1650 Max-Q de 4 Go semble être une mauvaise blague. En général, c’en est une.</p> <p>Sauf que le modèle testé est un MoE — un Mixture of Experts — et qu’un MoE ne calcule pas avec l’intégralité de ses paramètres à chaq…