# @jmbollenbacher — 2026-08-26

♥4 ↻1 · https://x.com/jmbollenbacher/status/2092640157849551101

After seeing the architectures glm-5.3-flash and qwen-38-flash-next, i think maybe this take was wrong.

We're going to push model sparsity even further. Larger models will require less compute and bandwidth in future gen models.

512GB could be worth having somday soon.

tags: author:jmbollenbacher, kind:tweet, on:glm-5-3-flash, on:qwen-3-8-flash-next, year:2026
cited on: _dossiers/glm-5-3-flash.md, _dossiers/qwen-3-8-flash-next.md, glm-5-3-flash, qwen-3-8-flash-next