HARDWARE FIT CHECK • Qwen 3.8 235B-A22B MoE + AMD Strix Halo Mini-PC (Ryzen AI Max+ 395 128GB)149.84 GB / 110 GB VRAM

Can I Run Qwen 3.8 235B-A22B MoE on AMD Strix Halo Mini-PC (Ryzen AI Max+ 395 128GB)?

Full VRAM footprint analysis, tokens/sec generation benchmark, and copy-paste Ollama & vLLM runtime commands for running Qwen 3.8 235B-A22B MoE (235B) on AMD Strix Halo Mini-PC (Ryzen AI Max+ 395 128GB) (110 GB).

✕ OOM149.84 GB (136%)0 t/s

Qwen 3.8 235B-A22B MoE on AMD Strix Halo Mini-PC (Ryzen AI Max+ 395 128GB)MoE (22B Active / 235B Total)

✕ Out of Memory — Exceeds VRAM Capacity

✕ OOM

Required VRAM

149.84 GB

136% of hardware capacity (110 GB)

Est. Decode Speed

0 tok/s

22B active params @ 256 GB/s

Available Headroom

-39.84 GB

Needs +39.84 GB more VRAM

Memory Allocation & Headroom

149.84 GB allocated of 110 GB capacity

136% utilized

Weights: 141 GBMoE (22B Active)

KV: (32K): 3.15 GB

Overhead: 5.69 GB

Deficit: -39.84 GB

Architecture & Throughput Transparency: Single-user decode speed (~0 tok/s) is memory-bandwidth bound at batch size = 1 on AMD Strix Halo Mini-PC (Ryzen AI Max+ 395 128GB) (256 GB/s). Prompt prefill / TTFT (~0 tok/s) is compute-bound and utilizes GPU Tensor cores. Actual latency varies with FlashAttention-2 and runtime engine (Ollama/vLLM/llama.cpp).

MoE Memory vs Speed Architecture: While decode throughput scales with the active parameter count (22B parameters streamed per token), all 235B parameters must be resident in VRAM so the dynamic router can route tokens to any expert layer.

Local Hardware Configurations & Pricing

Out of VRAM: High-Capacity Hardware Upgrades

High-bandwidth VRAM specs and current market prices

Multi-GPU Rig

48 GB VRAM

2x NVIDIA RTX 3090 Dual Rig (48GB Total VRAM)

Best price-to-VRAM homelab rig for Llama 3.3 70B Q4_K_M & Qwen 2.5 Coder 32B Q8_0

1648 GB/s700W TDP

$1,550Amazon Prime →

Apple Silicon

54 GB VRAM

Apple Mac Mini M4 Pro (64GB Unified Memory)

Silent 65W desktop running 32B–70B models with large KV context

273 GB/s65W TDP

$1,999Amazon Prime →

Apple Silicon↑ Upgrade

112 GB VRAM

Apple Mac Studio M4 Max (128GB Unified Memory)

Running Mistral Large 2 123B & Llama 3.3 70B Q8_0 locally with 128K context

546 GB/s145W TDP

$3,499Amazon Prime →

Unified APU / Mini-PC

110 GB VRAM

AMD Strix Halo Mini-PC (Ryzen AI Max+ 395 128GB)

Cheapest 128GB unified memory box for loading 70B–109B models

256 GB/s120W TDP

$1,899Amazon Prime →

Out of VRAM? Don't downgrade model quality.

Out of VRAM

Spin up an on-demand cloud GPU instance in 60 seconds with pre-installed Ollama and vLLM. (Qwen 3.8 235B-A22B MoE @ 149.84 GB VRAM)

RunPod Cloud GPUsBest 1-Click Ollama & vLLM

$0.38/hr (RTX 4090 Spot) – $2.39/hr (H100)

Launch 1-Click GPU Pod →

Vast.ai GPU MarketplaceCheapest Dual RTX 3090 / 4090

$0.18/hr (RTX 3060) – $0.78/hr (2x RTX 3090)

Rent Spot GPU →

Lambda GPU CloudEnterprise 8x H100 Clusters

$2.49/hr (H100) – $19.92/hr (8x H100)

Check H100 Availability →

Quantization Precision Matrix

32K ctx • KV: Q8_0

Quantization	Weights	KV Cache	Total VRAM	Est. Speed	Fit Status
FP1616 bpw	470 GB	3.15 GB	490.35 GB (446%)	0 t/s	OOM
Q8_08.5 bpw	249.1 GB	3.15 GB	261.72 GB (238%)	0 t/s	OOM
Q6_K6.56 bpw	192.7 GB	3.15 GB	203.34 GB (185%)	0 t/s	OOM
Q5_K_M5.68 bpw	166.85 GB	3.15 GB	176.59 GB (161%)	0 t/s	OOM
Q4_K_M4.8 bpw	141 GB	3.15 GB	149.84 GB (136%)	0 t/s	OOM
FP44.25 bpw	124.55 GB	3.15 GB	132.81 GB (121%)	~3 t/s	PARTIAL
IQ3_XS3.3 bpw	96.35 GB	3.15 GB	103.62 GB (94%)	~15 t/s	FITS

Export Full Architecture & Infrastructure Spec

Download an instant Markdown architecture spec with exact memory math, hardware benchmarks, and runtime CLI flags.

FAQ & Reference

Frequently Asked Questions: Qwen 3.8 235B-A22B MoE on AMD Strix Halo Mini-PC (Ryzen AI Max+ 395 128GB)

How much VRAM does Qwen 3.8 235B-A22B MoE need on AMD Strix Halo Mini-PC (Ryzen AI Max+ 395 128GB)?

At standard 4-bit quantization (Q4_K_M) with 32K context, Qwen 3.8 235B-A22B MoE requires 149.84 GB of total VRAM (weights + KV cache + CUDA runtime overhead).

What inference speed will Qwen 3.8 235B-A22B MoE achieve on AMD Strix Halo Mini-PC (Ryzen AI Max+ 395 128GB)?

Based on AMD Strix Halo Mini-PC (Ryzen AI Max+ 395 128GB)'s 256 GB/s memory bandwidth, expected decode speed is approximately 0 tokens per second.

Can I expand context beyond 32K tokens?

Each additional 32K tokens of context requires additional KV cache memory. For longer context windows, consider 8-bit or 4-bit KV cache quantization.