<!-- mobian-agent-page publisher="dailydev" canonical="https://daily.dev/posts/nvidia-dgx-spark-developer-kit-turns-a-desk-into-a-supercomputing-lab-gq0lvmti6" -->

---
title: NVIDIA DGX Spark Developer Kit Turns a Desk Into a...
description: NVIDIA has introduced the DGX Spark HPC Developer Kit, a four-node cluster of DGX Spark systems for researchers and developers. The nodes connect directly to...
canonical: https://daily.dev/posts/nvidia-dgx-spark-developer-kit-turns-a-desk-into-a-supercomputing-lab-gq0lvmti6
twitter:card: summary_large_image
twitter:site: @dailydotdev
og:type: website
og:site_name: daily.dev
og:title: NVIDIA DGX Spark Developer Kit Turns a Desk Into a Supercomputing Lab | daily.dev
og:description: NVIDIA has introduced the DGX Spark HPC Developer Kit, a four-node cluster of DGX Spark systems for researchers and developers. The nodes connect directly to...
og:url: https://daily.dev/posts/nvidia-dgx-spark-developer-kit-turns-a-desk-into-a-supercomputing-lab-gq0lvmti6
og:image: https://api.daily.dev/og/posts/GQ0lVMTI6.png
og:image:alt: NVIDIA DGX Spark Developer Kit Turns a Desk Into a Supercomputing Lab
og:image:width: 1200
og:image:height: 630
og:locale: en
---

> ## Documentation Index
> Fetch the complete documentation index at: https://daily.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# NVIDIA DGX Spark Developer Kit Turns a Desk Into a Supercomputing Lab

**[NVIDIA](https://daily.dev/sources/nvidia)** · 5 min read · 2 upvotes · 0 comments

## Summary

NVIDIA has introduced the DGX Spark HPC Developer Kit, a four-node cluster of DGX Spark systems for researchers and developers. The nodes connect directly to each other in a ring over ConnectX-7 networking, with no switch required. Combined, four nodes provide 512GB of memory and 20TB of NVMe storage, and with NVFP4 quantization can run models up to 700B parameters, such as DeepSeek V4.1. Demonstrated workloads include LAMMPS with Kokkos and NAMD, and the Rubin Lab at UC Berkeley reports results arriving 15x sooner on DGX Spark than with cluster queues. The kit is on display at SC26 and available for purchase through OEM partners and NVIDIA Marketplace.

## Full article

daily.dev links to this article rather than hosting it. Read it at the original source: <https://blogs.nvidia.com/blog/dgx-spark-hpc-developer-kit-sc26>

## Questions this post answers

### How many nodes can you link together in the NVIDIA DGX Spark HPC Developer Kit and what's the combined memory?

The DGX Spark HPC Developer Kit scales from one to four nodes, linked directly to each other in a ring over NVIDIA ConnectX-7 networking with no switch needed. At four nodes it provides 512GB of combined memory and 20TB of NVMe storage, and with NVFP4 quantization can run models with up to 700B parameters, such as DeepSeek V4.1, locally.

_Developers evaluating desktop HPC clusters can follow DGX Spark coverage and specs on daily.dev._

### Can DGX Spark run large models like DeepSeek V4.1 locally without a data center GPU cluster?

Yes, a four-node DGX Spark cluster can deploy models with up to 700B parameters locally using NVFP4 quantization, including DeepSeek V4.1. The Hartwig Group at UC Berkeley similarly fit a 120B-class Nemotron 3 Super model on a single DGX Spark using NVFP4, running continuous 8 to 24-hour training fits without multi-day cluster queue waits.

_Teams weighing on-prem versus cloud GPU options for large model deployment can track hardware news like this on daily.dev._

## Similar posts on daily.dev

- [Nvidia DGX Spark: The New Stack Developer’s Guide](https://daily.dev/posts/nvidia-dgx-spark-the-new-stack-developer-s-guide-xk9enevdy) · The New Stack · 2 upvotes · 0 comments
- [NVIDIA DGX Spark Powers Big Projects in Higher Education](https://daily.dev/posts/nvidia-dgx-spark-powers-big-projects-in-higher-education-rlb3io6wx) · NVIDIA · 0 upvotes · 0 comments
- [Dell's version of the DGX Spark fixes pain points](https://daily.dev/posts/dell-s-version-of-the-dgx-spark-fixes-pain-points-vy9zovd87) · Jeff Geerling · 7 upvotes · 1 comments
- [NVIDIA DGX Spark 64GB Gives Developers More Ways to Build and Scale Local AI](https://daily.dev/posts/nvidia-dgx-spark-64gb-gives-developers-more-ways-to-build-and-scale-local-ai-nmafldpqb) · NVIDIA · 1 upvotes · 0 comments
- [Scaling Autonomous AI Agents and Workloads with NVIDIA DGX Spark](https://daily.dev/posts/scaling-autonomous-ai-agents-and-workloads-with-nvidia-dgx-spark-lxoujr36b) · NVIDIA Developer · 1 upvotes · 0 comments

---

Tags: [#llm](https://daily.dev/tags/llm), [#gpu](https://daily.dev/tags/gpu), [#computing](https://daily.dev/tags/computing)

[View this post on daily.dev](https://daily.dev/posts/nvidia-dgx-spark-developer-kit-turns-a-desk-into-a-supercomputing-lab-gq0lvmti6)

```json
{"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]}
{"@context":"https://schema.org","@type":"TechArticle","headline":"NVIDIA DGX Spark Developer Kit Turns a Desk Into a Supercomputing Lab","url":"https://daily.dev/posts/nvidia-dgx-spark-developer-kit-turns-a-desk-into-a-supercomputing-lab-gq0lvmti6","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/nvidia-dgx-spark-developer-kit-turns-a-desk-into-a-supercomputing-lab-gq0lvmti6"},"datePublished":"2026-11-17T01:20:43.338Z","dateModified":"2026-11-17T01:23:08.717Z","description":"NVIDIA has introduced the DGX Spark HPC Developer Kit, a four-node cluster of DGX Spark systems for researchers and developers. The nodes connect directly to...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/575c894c0c2823eeadda77f7f62fe2d8?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/575c894c0c2823eeadda77f7f62fe2d8?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"NVIDIA","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"NVIDIA","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/236a5c1f410e450b8d696cf3a4bbe374","url":"https://daily.dev/sources/nvidia"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/nvidia-dgx-spark-developer-kit-turns-a-desk-into-a-supercomputing-lab-gq0lvmti6","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":2},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"llm,gpu,computing","timeRequired":"PT5M"}
{"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"NVIDIA","item":"https://daily.dev/sources/nvidia"},{"@type":"ListItem","position":3,"name":"NVIDIA DGX Spark Developer Kit Turns a Desk Into a Supercomputing Lab"}]}
{"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/nvidia-dgx-spark-developer-kit-turns-a-desk-into-a-supercomputing-lab-gq0lvmti6#faq","mainEntity":[{"@type":"Question","name":"How many nodes can you link together in the NVIDIA DGX Spark HPC Developer Kit and what's the combined memory?","acceptedAnswer":{"@type":"Answer","text":"The DGX Spark HPC Developer Kit scales from one to four nodes, linked directly to each other in a ring over NVIDIA ConnectX-7 networking with no switch needed. At four nodes it provides 512GB of combined memory and 20TB of NVMe storage, and with NVFP4 quantization can run models with up to 700B parameters, such as DeepSeek V4.1, locally. Developers evaluating desktop HPC clusters can follow DGX Spark coverage and specs on daily.dev."}},{"@type":"Question","name":"Can DGX Spark run large models like DeepSeek V4.1 locally without a data center GPU cluster?","acceptedAnswer":{"@type":"Answer","text":"Yes, a four-node DGX Spark cluster can deploy models with up to 700B parameters locally using NVFP4 quantization, including DeepSeek V4.1. The Hartwig Group at UC Berkeley similarly fit a 120B-class Nemotron 3 Super model on a single DGX Spark using NVFP4, running continuous 8 to 24-hour training fits without multi-day cluster queue waits. Teams weighing on-prem versus cloud GPU options for large model deployment can track hardware news like this on daily.dev."}}]}
```

