What Can You Do with DGX Spark? I Organized the Official Playbook
This page has been translated by machine translation. View original
Introduction
Hello, I'm Morishige from Classmethod's Manufacturing Business Technology Department.
I've been writing a DGX Spark series, from the unboxing to running Claude Code locally, but I still don't really understand "what can you actually do with DGX Spark?" (It's completely going to waste...)
As it turns out, NVIDIA officially publishes a collection of over 30 guides called "DGX Spark Playbooks."
So I organized all the Playbooks by category and created my own catalog. I hope it serves as a useful guide to the overall picture.
What Are Playbooks?
DGX Spark Playbooks are a collection of step-by-step guides published by NVIDIA in their official GitHub repository. They cover a wide range of topics, from inference engines to fine-tuning and robotics simulation.
Published under the Apache-2.0 license, community contributions are also welcome. Most Playbooks are configured to use NGC containers or Docker, so the barrier to setting up an environment is relatively low. You can access each Playbook from the official documentation site build.nvidia.com/spark.
DGX Spark was announced at CES 2025, and Playbooks began to be published around the time of its release in October 2025. Since then, CES 2026 announced up to 2.6x speedups through NVFP4 quantization and speculative decoding, and the Playbooks have been expanded to over 30.
Category-by-Category Catalog
Here I introduce all Playbooks confirmed as of February 2026, divided into 6 categories. Time estimates are based on the official site. Difficulty levels reflect the author's subjective assessment. The stability level (Stable / Beta / Experimental) of each Playbook is summarized in the "Stability Map" section later.
Inference Engines
This is the most well-developed category for DGX Spark. I felt it when I actually ran Ollama in the unboxing article, but the ability to run even gpt-oss:120b class models with 128GB of unified memory is truly powerful. With prefill at 282 tok/s and decode at 39 tok/s, the experience felt about as fast as reading speed.
| Playbook | Difficulty | Time Required | Summary |
|---|---|---|---|
| Ollama | Beginner | 10-15 min | Highly stable. Developer's first choice |
| Open WebUI with Ollama | Beginner | 10-15 min | Instant chat from browser. Ideal for onboarding |
| TRT-LLM for Inference | Intermediate | 45-60 min | Low-latency focused. NVFP4 quantization for 8B to 120B |
| vLLM for Inference | Intermediate | 30 min | For high-throughput and batch inference |
| SGLang for Inference | Intermediate | 30-45 min | Frontend language integration. Awaiting full SM121 support |
| NIM on Spark | Beginner | 15-30 min | Select from NGC catalog and deploy |
| Speculative Decoding | Intermediate | 30-45 min | EAGLE-3 accelerates gpt-oss:120B by 2-3x |
| Nemotron-3-Nano with llama.cpp | Beginner-Int. | 30 min | Run Nemotron 3 Nano 30B-A3B with llama.cpp |
Ollama is recommended as the developer's first choice for its ease of setup and stability.
TRT-LLM and vLLM have official NGC containers available and support a wide range of models. On the other hand, SGLang's support for the SM121 architecture is not yet complete, so it holds a Beta-like position.
Fine-Tuning
This is where DGX Spark's 128GB memory truly shines. Connecting two units via ConnectX-7 gives a 256GB configuration, making QLoRA fine-tuning of 70B models within reach.
| Playbook | Difficulty | Time Required | Summary |
|---|---|---|---|
| LLaMA Factory | Beginner-Int. | 30-60 min | Rapid experimentation with WebUI. Ideal for beginners |
| Fine-tune with PyTorch | Int.-Advanced | 1 hour | FSDP+LoRA supports 70B QLoRA with 2 Sparks |
| Fine-tune with NeMo | Advanced | 45-90 min | Production-scale distributed training |
| Unsloth on DGX Spark | Intermediate | 30-60 min | 2x speedup, memory efficient |
| FLUX.1 Dreambooth LoRA | Intermediate | 30 min + 1-2h train | FT of image generation model. Custom container |
| Vision-Language Model FT | Intermediate | 15-20 min + 1-2h train | Supports Qwen2.5-VL, InternVL3 |
For an easy start, LLaMA Factory is recommended. You can switch between SFT/LoRA/QLoRA/RLHF from the WebUI, making it well-suited as an introduction to fine-tuning. For a more serious approach, challenging 70B QLoRA with a two-unit PyTorch FT configuration sounds interesting.
Infrastructure & Connectivity
This is a group of Playbooks for getting DGX Spark into a "usable state." I think this category contains many things you'd set up first, like remote access and development environment setup with VS Code.
| Playbook | Difficulty | Time Required | Summary |
|---|---|---|---|
| Set Up Local Network Access | Beginner | 5 min | Local network configuration |
| Set up Tailscale | Beginner | 15-30 min | Secure remote access via VPN |
| DGX Dashboard | Beginner | 15-30 min | System monitoring and JupyterLab launch |
| VS Code | Beginner | 5 min | Remote development environment |
| Vibe Coding in VS Code | Beginner | 15-30 min | Continue.dev, Nsight CUDA Copilot |
| Connect Two Sparks | Intermediate | 1 hour | Connect 2 units at 200Gbps via ConnectX-7 |
| NCCL for Two Sparks | Intermediate | 30 min | Communication settings for distributed training |
Having Tailscale set up makes it very convenient to SSH in from anywhere. The Vibe Coding Playbook introduces Continue.dev and Nsight CUDA Copilot, allowing you to build a code completion environment using a local LLM.
Multimodal
There are only three, but this category features some visually impressive Playbooks.
| Playbook | Difficulty | Time Required | Summary |
|---|---|---|---|
| ComfyUI | Beginner | 30-45 min | Stable Diffusion runtime environment |
| Live VLM WebUI | Beginner-Int. | 20 min | Webcam → VLM real-time analysis |
| Multi-modal Inference | Intermediate | 1 hour | Multimodal inference pipeline |
Live VLM WebUI is one of the Playbooks I'm personally interested in. It can analyze webcam footage in real time with a VLM (Gemma 3 or Llama Vision) and display the results in a browser. Since it visualizes "what the AI is seeing," the impact as a demo is significant.
Applied Applications
These are Playbooks for building applications using LLMs, from RAG to knowledge graphs and video analysis.
| Playbook | Difficulty | Time Required | Summary |
|---|---|---|---|
| RAG in AI Workbench | Beginner | 30-45 min | Introduction to Retrieval-Augmented Generation |
| NVFP4 Quantization | Int.-Advanced | 45-90 min | NVFP4 quantization with TensorRT ModelOpt |
| Text to Knowledge Graph | Intermediate | 30 min | Visualization with Ollama+ArangoDB+Three.js |
| Video Search & Summarization | Int.-Advanced | 30-45 min | VLM+LLM+RAG integrated video analysis |
| Multi-Agent Chatbot | Intermediate | 1 hour | Multi-agent system |
Text to Knowledge Graph uses Ollama to extract triples (subject-predicate-object) from unstructured text and interactively visualizes them with ArangoDB+Three.js. With 128GB of memory, you can perform high-accuracy triple extraction with a 70B model, making it a use case where DGX Spark's unified memory really shines.
Other (Robotics, Data Science, Domain-Specific)
This is a category of niche but interesting Playbooks.
| Playbook | Difficulty | Time Required | Summary |
|---|---|---|---|
| Isaac Sim / Isaac Lab | Int.-Advanced | 30 min + 10-15 min build | GPU-accelerated physics simulation & reinforcement learning |
| CUDA-X Data Science | Beginner | 20-30 min | GPU acceleration with RAPIDS, cuML, cuDF |
| Optimized JAX | Intermediate | 2-3 hours | Optimized JAX development environment |
| Single-cell RNA Sequencing | Advanced | 15 min | Genomics workflow |
| Quantitative Portfolio Optimization | Advanced | 20 min | Financial quantitative analysis |
Isaac Sim / Isaac Lab is a noteworthy Playbook for robotics developers. It integrates PhysX-based GPU-accelerated physics simulation and an RL (reinforcement learning) framework, with over 30 preset learning environments for robots like ANYmal, Unitree Go2, and Franka Panda. A Sim2Real pipeline becomes visible: simulate on DGX Spark → deploy on real hardware with Jetson. The community has also reported successful Sim2Real transfer cases, including the Disney BDX droid, low-cost autonomous racing, and quadruped robot door-opening, making this a very interesting direction.
The domain-specific Playbooks (genomics, finance) are quite specialized in content, with a limited target audience. However, it's interesting that they demonstrate the possibility of using DGX Spark in these specialized fields as well.
Stability Map
The following is my own personal interpretation, but I have classified all Playbooks into 3 levels of stability.
| Stability | Definition | Examples |
|---|---|---|
| Stable | Official NGC container provided, verified with multiple models, extensive troubleshooting available | Ollama, TRT-LLM, vLLM, PyTorch FT, NeMo FT, LLaMA Factory |
| Beta | Works but marked "Beta," or DGX Spark-specific caveats remain | NIM, SGLang, Speculative Decoding, NVFP4, Live VLM WebUI, Nemotron-3-Nano |
| Experimental | Uses custom container, verified only with specific models | FLUX FT, VLM FT, Multi-Agent, Isaac Sim, Knowledge Graph |
SM121 Architecture Challenges
DGX Spark's GB10 chip adopts a new architecture called SM121, and software support has not yet fully caught up.
- Triton (GPU compiler) does not yet support SM121
- Reports of CUDA errors in some FP8 quantization kernels (BF16 recommended for fine-tuning)
- Full SM121 support for SGLang and vLLM is still a work in progress
NVIDIA announced a 2.6x performance improvement over FP8 via NVFP4 + speculative decoding at CES 2026, and community-based support is also active. However, for the time being, I think it's safest to "start with Stable items and gradually expand from there."
Specifically, the areas most likely to run into SM121 issues are inference with vLLM/SGLang and fine-tuning with FP8. Inference with Ollama and fine-tuning with BF16 work stably.
Where to Start
With over 30 Playbooks to choose from, I think many people will be unsure where to begin, so here are recommended paths for 3 levels.
Start Here (Within 30 Minutes)
Set up Ollama + Open WebUI. You can pull gpt-oss:20b and reach a state where you can chat from a browser in 10-15 minutes. For detailed steps, please refer to the unboxing article mentioned earlier.
Go a Little Deeper
From here, there are two directions.
If you want to pursue inference speed, the TRT-LLM Playbook is recommended. You can experience the speed difference compared to Ollama firsthand, and using NVFP4 quantization, community reports suggest figures around 5.39 tok/s for Llama 3.3 70B.
If you're interested in fine-tuning, starting with LLaMA Factory is the easiest approach. You can adjust parameters from the WebUI and try SFT/LoRA/QLoRA with your own dataset.
For Serious Use
This is the 256GB configuration of two DGX Sparks connected via ConnectX-7. This configuration enables the following workloads:
- 70B QLoRA fine-tuning (FSDP+LoRA, PyTorch FT Playbook)
- NVFP4 inference of Qwen3-235B-A22B (community reports suggest around 11.73 tok/s)
- 2-3x speedup with EAGLE-3 speculative decoding
I haven't been able to try this area myself yet, so I'll report back once I do.
Summary
I organized the overall picture of DGX Spark Playbooks by category. While challenges remain in the software maturity of the SM121 architecture, you can use it stably by starting with Stable Playbooks like Ollama and TRT-LLM, and fine-tuning with BF16 is also practical. The ecosystem continues to expand, with NVFP4 optimizations at CES 2026 and the addition of Nemotron 3 Nano.
I recommend first experiencing the benefits of 128GB memory with Ollama + Open WebUI, then expanding in the direction that interests you. I hope this article serves as a useful reference for choosing Playbooks.
DGX Spark Series?
| # | Title |
|---|---|
| Article 1 | DGX Spark Has Arrived |
| Article 1.5 | Retrying Local Execution of Claude Code with DGX Spark's 128GB Memory |
| Bonus | What Can You Do with DGX Spark? Organizing the Official Playbooks (This Article) |
Reference Links
Official Resources
- DGX Spark Playbooks GitHub
- DGX Spark Playbooks Documentation
- DGX Spark Developer Forum
- DGX Spark Release Notes
- Isaac Sim / Isaac Lab
Related Technical Articles
- New Software and Model Optimizations Supercharge DGX Spark (CES 2026)
- Level1Techs DGX Spark Review
- Simon Willison DGX Spark Review
