# Post Event Q&A - PyTorch Afters: Efficient World Model and LLM Training on Blackwell Infrastructure

**URL:** https://forum.verda.com/t/post-event-q-a-pytorch-afters-efficient-world-model-and-llm-training-on-blackwell-infrastructure/69
**Category:** Event
**Created:** [October 21, 2025, 9:42pm UTC](https://forum.verda.com/t/post-event-q-a-pytorch-afters-efficient-world-model-and-llm-training-on-blackwell-infrastructure/69 "2025-10-21T21:42:46Z")
**Posts on this page:** 1
**Page:** 1

<div class="post-metadata">

### Author: ![TeamVerda](https://dub1.discourse-cdn.com/flex014/user_avatar/forum.verda.com/teamverda/32/36_2.png) [@TeamVerda](https://forum.verda.com/u/TeamVerda)
#### Post date: [October 21, 2025, 9:42pm UTC](https://forum.verda.com/t/post-event-q-a-pytorch-afters-efficient-world-model-and-llm-training-on-blackwell-infrastructure/69/1 "2025-10-21T21:42:46Z")

</div>

**What did we cover?**  
We explored how next-generation **AI infrastructure** is shaping the future of **reinforcement learning, large-scale inference, and world models**.

**Who spoke?**

- **Paul Chang (DataCrunch):** _Training World Models Using B200s_

- **Erik Schultheis (IST Austria):** _Training Quantized LLMs Efficiently on Consumer GPUs_

💭 **Any questions, thoughts, or takeaways from the session?**  
Drop them below ⬇ we’d love to hear your perspective and keep the discussion going.