# NVIDIA Shows 3 Ways to Speed Up MoE Inference on DGX Spark, Lifting Ling-3.0-flash Throughput 42.5%

> **Publication**: aitalky (Independent AI Journalism)
> **Canonical Citation URL**: https://aitalky.vercel.app/news/nvidia-shows-3-ways-to-speed-up-moe-inference-on-dgx-spark-lifting-ling-3-0-flash
> **Original Source**: [MarkTechPost](https://www.marktechpost.com/2026/11/03/nvidia-dgx-spark-moe-inference-ling-3-0-flash/)
> **Author**: Asif Razzaq (undefined)
> **Published Date**: Tuesday, November 3, 2026
> **Category**: RESEARCH
> **Tags**: AI, MarkTechPost, research

## Executive Summary
NVIDIA shows how 3 stacked optimizations speed up MoE inference on a single DGX Spark. On Ant Group inclusionAI's 124B Ling-3.0-flash, GSM8K decode throughput rose 42.5% with no detectable accuracy change



## Full Coverage
NVIDIA shows how 3 stacked optimizations speed up MoE inference on a single DGX Spark. On Ant Group inclusionAI's 124B Ling-3.0-flash, GSM8K decode throughput rose 42.5% with no detectable accuracy change

---
*Generated for AI agents and LLM citation by aitalky (https://aitalky.vercel.app).*