NVIDIA DGX Spark
The power of NVIDIA Grace
Blackwell on your desk.
About
NVIDIA DGX SparkTM is a compact, power-efficient personal AI supercomputer powered by the NVIDIA GB10 Grace Blackwell Superchip. Delivering up to one petaFLOP of FP4 AI performance and 128 GB of unified memory, it puts the power to prototype, fine-tune, and deploy the latest AI models directly on a desktop machine — no data centre required.
With the full NVIDIA AI stack built in and NVIDIA NemoClaw enabling autonomous agent development from the desktop, developers can move from experimentation to real-world impact faster than ever.
This is what we are bringing to you — through the RP Tech x NVIDIA DGX Spark Developer Meetup in Bangalore on April 11, and an online webinar — featuring live demos and developer-focused workflows designed to help you build smarter and ship faster.
GenAI at Your Desk: Building Powerful AI with NVIDIA DGX Spark
Building generative AI applications often requires significant compute power and complex development workflows. But what if you could bring that capability directly to your desktop?In this session, discover how NVIDIA DGX Spark enables developers to prototype, fine-tune, and run large AI models locally—making experimentation faster, more efficient, and highly scalable.
Leverage the full NVIDIA AI stack—including CUDA, TensorRT, NeMo, and NIM microservices —to seamlessly build, optimize, and deploy GenAI applications. From rapid prototyping to real-world deployment, learn how to streamline your development journey without compromising on performance.
Amit Kumar
Manager Solutions Architecture and Engineering | Industries | GCCS, NVIDIA
Arsh Goyal
AI & Engineering Influencer
Saturday, April 11, 2026
5.30 pm Onwards
Hyatt Centric, MG Road, Bangalore
Agenda
6:15 PM
Keynote by Arsh Goyal, AI & Engineering Influencer
6:25 PM
Introduction to NVIDIA DGX Spark by Amit Kumar, Manager Solutions Architecture and Engineering | Industries | GCCS, NVIDIA
6:55 PM
NVIDIA DGX Spark Demo: Your personal AI supercomputer for agentic LLMs. Train, tune, and deploy agents, right on your desk.
8:25 PM
Networking & Dinner
Webinar
Friday, April 17, 2026
3.00 - 4.30 pm
Running Sovereign LLMs on NVIDIA DGX Spark (Hands-on)
In this webinar, we will walk you through how to run sovereign LLM models on NVIDIA DGX Spark, bringing powerful AI capabilities directly to your desktop environment.
Watch a live demonstration of a local chat application powered by advanced models such as Sarvam 30B and Param2-17B, showcasing how developers can efficiently deploy and interact with large language models in a secure, local setup.
Learn how to leverage DGX Spark to run high-performance AI workloads, enabling faster experimentation, enhanced data privacy, and seamless development workflows.
Megh Makwana
Manager - Applied GenAl Solution Engineering, NVIDIA
Session Flow
Welcome and Introduction
Introduction to NVIDIA DGX Spark by Megh Makwana
Hands-on: Demonstration of running Sovereign LLM like Sarvam 30B and Param-2-17B on NVIDIA DGX Spark
Open House (Q&A)
Key Takeaways
Copyright © 2026 NVIDIA Corporation
