text model · LFM · Windows
Can I run LFM2 350M on AMD Ryzen AI Halo (128GB)?
Yes. LFM2 350M runs on AMD Ryzen AI Halo (128GB) at Q4_K_M (~1.4 GB of ~96 GB usable).
Runs at Q4_K_M using ~1.4 GB of ~96 GB usable. You have room for FP16 for higher quality.
That figure is at a 4k context and moves about ±15% as context length changes. AMD Ryzen AI Halo (128GB) leaves ~94.6 GB of headroom, room to step up to FP16 for higher quality.
- Q4_K_M needed
- ~1.4 GB
- Usable on device
- ~96 GB
- Device memory
- 128 GB
- Best quant
- Q4_K_M
Which quant fits
Running cost · estimate
- Power draw
- ~120 W
- Electricity / 1M tokens
- ~$0.01
- Pays for itself after
- ~8,161M tok
At ~$0.15/kWh and the estimated ~370 tok/s, a million generated tokens costs about $0.01 in electricity. Versus a hosted API at ~$0.5 per million tokens, the ~$3,999 AMD Ryzen AI Halo (128GB) pays for itself after roughly 8,161 million tokens, so local hardware is mostly a fixed cost, not a per-token one. TDP is the peak draw, so this is an upper bound. Assumptions.
Run it
Pick your tool. All 2 load the same Q4_K_M weights.
llama-cli -hf LiquidAI/LFM2-350M-GGUF:Q4_K_M lms get LiquidAI/LFM2-350M-GGUF How to run it
On Windows use LM Studio (Best GUI on Windows, auto-detects CUDA/Vulkan backends.).
- Parameters
- 0.354B
- Q4_K_M size
- 0.45 GB
- Q8_0 size
- 0.35 GB
- Context
- 128k
- Memory
- 128 GB unified
- Usable for weights
- ~96 GB
- Power draw
- ~120 W
- Best runtime
- llama.cpp (Vulkan/ROCm) / LM Studio
You could also run
Run LFM2 350M on other hardware
FAQ
Can AMD Ryzen AI Halo (128GB) run LFM2 350M?
Yes. LFM2 350M runs on AMD Ryzen AI Halo (128GB) at Q4_K_M (~1.4 GB of ~96 GB usable).
How much memory does LFM2 350M need?
AMD Ryzen AI Halo (128GB) has room to spare. At Q4_K_M the weights are ~0.45 GB; with KV cache and runtime overhead, budget ~1.4 GB at a 4k context.
What is the best tool to run LFM2 350M on Windows?
LM Studio for a simple setup; Ollama (CUDA) for the most speed. AMD GPUs run via Vulkan/ROCm at roughly half CUDA throughput. NVIDIA is the smooth path on Windows.
Embed this
A live badge for your model card or README, updated as the data is. [](https://localmodel.run/can-i-run/lfm2-350m/amd-ryzen-ai-halo-128gb) Sources
Weights are measured from GGUF files; KV cache and overhead are computed, so totals can vary ~15% with context and runtime. Any tok/s is a bandwidth estimate. See methodology.