HN Showcase
Back to browse
GitHub Repository

High-performance interactive system monitor for NVIDIA DGX systems — GPU, CPU, memory, disk, network in a beautiful TUI

44 starsRust

Dgxtop: Rust Monitor for DGX or need to monitor GPU status

by dennysora-main·Jul 12, 2026·2 points·0 comments
Visit ProjectView on HN

AI Analysis

●●SolidNiche GemSlick

Direct NVML access beats nvidia-smi subprocess calls for speed and reliability.

Strengths
  • •Direct NVML library integration avoids the latency and parsing overhead of spawning nvidia-smi subprocesses.
  • •Interactive TUI allows sorting, filtering, and killing GPU processes directly without leaving the terminal.
  • •Tracks NVLink topology and ECC errors, which are critical for debugging multi-GPU training jobs.
Weaknesses
  • •Highly specialized for DGX hardware; standard workstations won't benefit from NVLink or ECC features.
  • •Competes with established tools like nvtop and btop that already cover general GPU monitoring needs.
Target Audience

ML engineers and sysadmins managing NVIDIA DGX clusters

Similar To

nvtop · btop · nvidia-smi

Post Description

This Project is make it easy to monitor the GPU infomation.

Any platform can work, If can't get the Nvidia GPU message, also can work. it will to get cpu, disk and network io information.

This Project will get detail infomation about the GPU.

And helps monitor the training process.

Similar Projects

SaaS●●Solid

SiteReady – Uptime monitoring and status pages for indie makers

Clean, focused product: email alerts, response-time tracking, and instant public status pages with a forever-free 2-monitor tier make it easy to get started. It isn't reinventing monitoring, but the stripped-down pricing and one-click status publishing feel deliberately built for solo founders — missing advanced alerting (SMS/pager/on-call tooling) and broader integrations keeps it squarely aimed at small shops.

Niche GemShip ItSlick
Arifcodes
114mo ago
Infrastructure●Mid

GPU-hot Dashboard for monitoring Nvidia GPUs on remote servers

Shoots for zero-setup GPU visibility: one docker run spins up a service you open in the browser to see live NVIDIA metrics without Prometheus, SSH, or dashboards to configure. The UI and interactive demo show attention to UX and make it instantly useful for small clusters or single-node setups. It doesn’t reinvent observability — if you need long-term metrics, alerting, or enterprise integrations you’ll still reach for exporters + Grafana — but for lightweight, immediate GPU troubleshooting this is convenient and focused.

Niche GemSlick
github-trending
214mo ago
Developer Tools●●●Banger

cuTile Rust: Safe, data-race-free GPU kernels in Rust

Extends Rust's ownership model across GPU boundary with tile-based partitioning for data-race-free kernels.

WizardryBig BrainNiche Gem
melihelibol
1061827d ago