Code, Test, and Deploy on One Machine with NVIDIA RTX Spark

Code, Test, and Deploy on One Machine with NVIDIA RTX Spark

The landscape of software development is shifting. Developers are moving away from writing every line of code manually toward directing agents that generate most of the implementation. Historically, this workflow has required a fragmented setup: a laptop for coding, a cloud model API for intelligence, and a cloud VM for deployment. NVIDIA RTX Spark aims to consolidate this entire process onto a single machine, bringing coding agents, model customization, and CUDA-accelerated development directly to the everyday PC.

One Machine to Code, Test, and Deploy

NVIDIA RTX Spark combines the NVIDIA GB10 system-on-a-chip with up to 128GB of unified memory. This architecture allows developers to run local coding agents, access AI frameworks, and deploy applications on a single PC without needing external cloud infrastructure for the core development loop.

The platform is designed to handle the full spectrum of AI development workloads. With up to 128GB of unified memory and 1 petaflop of FP4 AI compute, developers can run coding models locally, including Qwen 3.8 Flash Next, significantly reducing cloud coding agent token consumption. The hardware includes the NVIDIA Grace CPU and Blackwell GPU, providing the necessary compute for inference, fine-tuning, and compilation.

According to NVIDIA, RTX Spark delivers significant performance gains compared to Intel Panther Lake 12Xe. The platform offers 4.6x faster coding-agent performance, 2.7x faster fine-tuning, and 1.8x faster code compilation. This shift moves development from a CPU-bound workload, where performance was determined mostly by the CPU, to a GPU-bound workload where inference and testing are accelerated by dedicated hardware.

A Familiar Software Stack, Expanded for Windows

To support this hardware, NVIDIA partnered with Microsoft to expand the Windows on Arm software stack. This collaboration introduces several key tools for the developer community:

• PyTorch-CUDA Build: The release of the first PyTorch-CUDA build for this architecture.
• Python 3.15: A native Python release by default.
• Conda Packages: Over 2,000 ready-to-use Conda packages.
• Package Compatibility: Arm-compatible versions for more than 97% of the top 15,000 Python packages.

Developers can choose to work in Windows, including the new Windows Developer Config developed with Microsoft. This configuration installs top development frameworks, mutes notifications, enhances the terminal, and allows moving the task bar to the side. Alternatively, developers can work in Windows Subsystem for Linux (WSL), gaining access to the same environment and dependencies they will use when deploying to the cloud. This approach enables developers to complete all their work in their target OS without code migrations.

Since RTX Spark features an NVIDIA GPU, developers can test and deploy to the largest local AI installed base of 150 million RTX PCs, as well as virtually all clouds that run on NVIDIA GPUs.

Local Coding Agents Come to RTX Spark

RTX Spark devices are ideal for local coding agents. NVIDIA has collaborated closely with top agents, including GitHub Copilot, Pi, Hermes, and OpenClaw, to improve them and make them easier to use.

Security and Hybrid Routers

Enhanced security is provided through NVIDIA OpenShell and/or Microsoft Execution Containers, delivering OS-level sandboxing. Additionally, hybrid prompt routers allow prompts to be sent to local models if they are sufficient, or escalated to cloud models if needed. This reduces token consumption while maintaining capability.

Microsoft GitHub Copilot now enables seamless local AI through Windows ML, supported with llama.cpp as a backend. The prompt router HydraFusion has been upgraded to support local models. The new MAI 1.1 Coding Flash model can also run locally on RTX Spark.

Pi, an open-source coding agent with a lightweight terminal interface and support for local and cloud models, is optimized for RTX Spark. Its extensions, skills, and prompt templates give developers control over how the agent fits into their workflow, from model selection to the tools it can use.

JetBrains Junie Local

JetBrains has launched native support for Junie Local on RTX Spark. This coding agent reads code, edits files, and runs tests, keeping the model and code on the machine. Junie Local sizes its context window to the memory available to the GPU, allowing RTX Spark's unified memory to keep more code, tool output, and conversation history in context.

New Nemotron Models Offer More Options for Local Agents

Nemotron 3.5 Bolt 74B-A9B 24GB is a new large language model (LLM) purpose-built for running local agents on RTX Spark. It is a 74B-parameter mixture-of-experts (MoE) model with 9B active parameters. The model is packed into 24GB using 2-bit quantization with advanced quantization-aware training techniques to retain quality.

NVIDIA states that this model aims to achieve the optimal balance between speed and intelligence, being faster than most popular dense models while smarter than most popular MoE models with a similar memory footprint. Users will be able to run Nemotron 3.5 Bolt 74B at launch through llama.cpp, vLLM, Ollama, Unsloth, and LM Studio. It will also be directly accessible via agents such as Hermes and OpenClaw.

Like other Nemotron models, Bolt 74B is fully open source, including training data, weights, and recipes, so developers can fine-tune and customize their own agents.

Decision Models Now Supported in Top Frameworks

Decision models represent a breakthrough in AI development, giving agents a near-instant way to make decisions without lengthy reasoning loops. They can choose what to do next, from classifying requests to selecting the next action, up to 200x faster than traditional reasoning loops.

These decision models are now supported in llama.cpp and Unsloth on RTX Spark and GeForce RTX GPUs. This capability allows agents to perform rapid decision-making tasks efficiently on local hardware.

AI for Media SDKs

The NVIDIA AI for Media SDKs are now available for RTX Spark. These SDKs include support for video super resolution, frame generation, and TrueHDR. The Python wheel is available on PyPI, with C++ SDKs also available on NGC for RTX Spark and Windows on Arm.

Bring Your Windows Applications to RTX Spark

To bring an existing application to RTX Spark, developers should first assess dependencies for Windows on Arm. Options include choosing a native Arm64 build or using Arm64EC to migrate components incrementally. Developers are advised to test installation, functionality, performance, and the NVIDIA and CUDA paths their app uses.

NVIDIA provides app-porting resources that explain migration options and tools, including how to start preparation on an existing Windows on Arm development system.

Find Your RTX Spark and Start Building

RTX Spark is currently available in laptops and is coming in November to small form factor desktops, which are ideal for always-on, 24×7 running agents.

To get started, developers can follow the AI Development playbook to configure Windows, set up their agent, and connect NVIDIA Skills. NVIDIA also provides a repository of NVIDIA Skills that teach agents how to use CUDA-X libraries, enabling developers to build GPU-accelerated applications without requiring deep familiarity with CUDA.

Because NVIDIA GPUs are the most widely used platform for AI development, new tools and frameworks typically work on NVIDIA from day one, while support for other hardware often follows weeks later. There is also access to NVIDIA CUDA-X libraries, including tools such as FlashInfer for LLM acceleration and NVIDIA NeMo libraries for AI model development.

Explore RTX Spark systems to find available purchasing options and begin leveraging the power of local AI development.

Source: https://developer.nvidia.com/blog/code-test-and-deploy-on-one-machine-with-nvidia-rtx-spark