<p><span class="h-card" translate="no"><a href="https://awoo.space/@DPA" class="u-url mention">@<span>DPA</span></a></span> you do not need a supercomputer to run an LLM, and synthesis already relies on SAT in basically all major toolchains</p>