NVIDIA AI Retweets DGX Spark Qwen Performance Post
Official NVIDIA AI account retweets post on Qwen model runs across DGX Spark systems.
TLDR
The official NVIDIA AI account retweeted a post by WescheNex1q that described inference testing on DGX Spark hardware with the Qwen3.8 Flash Next model under NVFP4 quantization. The post referenced multiple concurrent users each running independent prompts with separate key-value caches. The retweet presents the original poster's description of the test but offers no additional verification or details from NVIDIA itself.
Combined views
813
1 Source, first seen ago