DigestAI news desk

Cut through the AI noise.

Model

Qwen3-30B-A3B news

The latest Qwen3-30B-A3B news, : Prime Intellect releases Prime Mega MoE kernel for faster training.

2 stories mentioning Qwen3-30B-A3B since 21 September 2026, newest first, each with its sources and discussion. Follow to see new ones on your front page.

  1. Research new

    Prime Intellect releases Prime Mega MoE kernel for faster training

    Prime Intellect has released Prime Mega MoE, a GPU kernel extension that adds a fused, overlapping backward pass to DeepSeek’s existing Mega MoE kernel. While DeepSeek’s original kernel only supported the forward pass…

    1 source
    MarkTechPost
  2. StepFun launches Step 5 Preview with 1M-token context and $1 input pricing

    StepFun announced Step 5 Preview, its flagship mixture‑of‑experts model aimed at agentic software engineering, knowledge work, and finance tasks. The model totals about 600 billion parameters but activates roughly 27…

    1 source
    MarkTechPost

Questions about Qwen3-30B-A3B

What is the latest Qwen3-30B-A3B news?

As of 29 October 2026: Prime Intellect releases Prime Mega MoE kernel for faster training. Prime Intellect has released Prime Mega MoE, a GPU kernel extension that adds a fused, overlapping backward pass to DeepSeek’s existing Mega MoE kernel.