# @storagereview on Instagram

- **Type:** Image
- **Original URL:** https://www.instagram.com/p/DYNqzo6GB_7
- **Gondola URL:** https://gondola.cc/posts/65393809-storagereview-instagram
- **Thumbnail:** https://img.gondola.cc/tr:w-,h-,fo-auto/postThumbnails/20eee05cc1.jpg
- **Posted:** 2026-05-11T21:32:20.000+00:00
- **Account Owner:** StorageReview (@storagereview) — https://gondola.cc/storagereview

## Caption

Two DGX Sparks. Three OEMs. One 200 GbE fabric and a methodological detour that produced the most interesting result of the review. We benchmarked dual-Spark clusters from Dell, GIGABYTE, and HP across multiple models and workload shapes, and found the OEMs landed within a narrow band. The real finding was about pipeline parallelism beating NVIDIA’s tensor parallelism default by 2.20x on GPT-OSS-120B once concurrency scaled. Link in bio.

#NVIDIA #DGXSpark #AI #inference

## Stats

- **Views:** 0
- **Likes:** 243
- **Shares:** 0
- **Comments:** 2

## Tags

inference, dgxspark, ai, nvidia

---
Copyright (c) Gondola