News for hackers. Stuff that matters.

A comment on NVIDIA DGX Spark In-Depth Review: A New Standard for Local AI Inference. Parent comment

Re:NVIDIA DGX Spark In-Depth Review: A New Standard for Loca... by KeplerBoy on Wednesday October 15, 2025 @03:34AM (#45589196)
The jetson thor seems to be quite different. The Thor whitepaper lists 8 TFlop/s of FP32 compute where the DGX sparks seems to be closer to 30 TFlop/s. Also 48 SMs on the Spark vs 20 on the Jetson.

The DGX seems vastly more capable.

[ Parent | on HN ]

There are no comments on this comment yet.