Just did my last upgrade of the year: added another 2×32GB RAM. My local LLM rig 4x100 96G VRAM + 192G RAM is now complete. 128GB was already plenty, but I’m prepping for new models (Qwen4-Flash?)—or maybe one day Strata makes Q6 as fast as Q4. Add 1M context to that? Absolutely insane.
- Just did my last upgrade of the year: added another 2×32GB RAM. My local LLM rig 4x100 96G VRAM + 192G RAM is now complete. 128GB was already plenty, but I’m prepping for new models (Qwen4-Flash?)—or maybe one day Strata makes Q6 as fast as Q4. Add 1M context to that? Absolutely insane.
prepping for models that don't exist yet is peak local llm. 192gb means you're ready the day qwen4 actually ships
Localmaxxing
Love the prep for bigger models. Still running Strata on a 2016 Dell R730 (V100 16GB + P100 12GB, ~153GB usable). Just added a second box with a 5060 Ti though, and yeah, RAM is the new bottleneck. You aiming Q6 first or the long context?