# Dual Spark Ducted Cooling Cage

**URL:** <https://forums.developer.nvidia.com/t/dual-spark-ducted-cooling-cage/365302>\
**Category:** DGX Spark / GB10\
**Created:** [April 1, 2026, 3:58am UTC](https://forums.developer.nvidia.com/t/dual-spark-ducted-cooling-cage/365302 "2026-04-01T03:58:34Z")\
**Posts on this page:** 1\
**Showing post:** 325

<div class="post-metadata">

**Author:** ![griffith.mark](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/griffith.mark/32/453455_2.png) [@griffith.mark](https://forums.developer.nvidia.com/u/griffith.mark)\
**Post date:** [August 19, 2026, 1:57am UTC](https://forums.developer.nvidia.com/t/dual-spark-ducted-cooling-cage/365302/325 "2026-08-19T01:57:30Z")

</div>

For anyone building ducts, cages, or fan mods in this thread: before you spend more time on hardware, read the software‑only approach here:

\*\*[Cooler GB10 Temps – (almost) no performance lost](https://forums.developer.nvidia.com/t/cooler-gb10-temps-almost-no-performance-lost/372662)  
\*\*

Key points:

- Capping the GPU clocks (e.g. `sudo nvidia-smi -lgc 0,2000`) gives essentially the same real‑world performance for many LLM/image workloads, with much lower power and temps.
- A simple closed‑loop governor that holds your hottest thermal zone at a target temperature turns any cooling headroom directly into extra clocks, so better cooling shows up as MHz, not just °C.

If you’re evaluating a duct or cage, use that governor + hottest‑zone logging and report “this duct is worth X MHz at a fixed thermal target” instead of just “temps are lower”.

---

_[View the full topic](https://forums.developer.nvidia.com/t/dual-spark-ducted-cooling-cage/365302)._