Skip to content
localmodel.run

text model · SmolLM2 · Linux

Can I run SmolLM2 360M on Nvidia DGX Spark (128GB)?

Compatibility verdict VRAM threshold engine
Yes, it runs GPU accelerated ~655 tok/s est.

Yes. SmolLM2 360M runs on Nvidia DGX Spark (128GB) at Q4_K_M (~1.2 GB of ~112 GB usable).

Needs ~1.2 GB Device usable ~112 GB

Runs at Q4_K_M using ~1.2 GB of ~112 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. Nvidia DGX Spark (128GB) leaves ~110.8 GB of headroom, room to step up to FP16 for higher quality.

Q4_K_M needed
~1.2 GB
Usable on device
~112 GB
Device memory
128 GB
Best quant
Q4_K_M
Share on X Share on Reddit

Which quant fits

Quant ladder vs ~112 GB usable
Q2_K
~1.1 GB
Q3_K_M
~1.1 GB
Q4_K_M
~1.2 GB
Q5_K_M
~1.2 GB
Q6_K
~1.2 GB
Q8_0
~1.3 GB
FP16
~1.6 GB
The line marks Nvidia DGX Spark (128GB)'s ~112 GB budget; rungs past it are too large.

Running cost · estimate

Power & economics est.
Power draw
~140 W
Electricity / 1M tokens
~$0.01
Pays for itself after
~8,161M tok

At ~$0.15/kWh and the estimated ~655 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 Nvidia DGX Spark (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

Install commands Linux

Pick your tool. All 3 load the same Q4_K_M weights.

Ollama
$ ollama run smollm2:360m
llama.cpp
$ llama-cli -hf bartowski/SmolLM2-360M-Instruct-GGUF:Q4_K_M
LM Studio
$ lms get bartowski/SmolLM2-360M-Instruct-GGUF
Model SmolLM2
Parameters
0.362B
Q4_K_M size
0.271 GB
Q8_0 size
0.386 GB
Context
2k
Ollama tag
smollm2:360m

Set up this model at 2K →

Full SmolLM2 360M requirements →
Device Linux
Memory
128 GB unified
Usable for weights
~112 GB
Power draw
~140 W
Best runtime
llama.cpp (CUDA) / Ollama (CUDA)
Best models for Nvidia DGX Spark (128GB) →

You could also run

Run SmolLM2 360M on other hardware

Nvidia GeForce RTX 2060 (6GB)Apple M1 (8GB)8GB RAM Laptop (CPU/iGPU only)iPhone 15 ProiPhone 16iPhone 16 Pro

FAQ

Can Nvidia DGX Spark (128GB) run SmolLM2 360M?

Yes. SmolLM2 360M runs on Nvidia DGX Spark (128GB) at Q4_K_M (~1.2 GB of ~112 GB usable).

How much memory does SmolLM2 360M need?

Nvidia DGX Spark (128GB) has room to spare. At Q4_K_M the weights are ~0.271 GB; with KV cache and runtime overhead, budget ~1.2 GB at a 4k context.

What is the best tool to run SmolLM2 360M on Linux?

Ollama for a simple setup; vLLM for the most speed. vLLM shines for multi-user serving/throughput. For a single local chat, Ollama or llama.cpp is simpler and lighter.

Embed this

SmolLM2 360M on Nvidia DGX Spark (128GB) compatibility badge A live badge for your model card or README, updated as the data is.
Markdown
$ [![SmolLM2 360M on Nvidia DGX Spark (128GB)](https://localmodel.run/badge/smollm2-360m/nvidia-dgx-spark-128gb.svg)](https://localmodel.run/can-i-run/smollm2-360m/nvidia-dgx-spark-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.