Back to AI intel
重点
Ollama v0.31.2-rc2 enables iGPU mmproj offload with fit padding.
AI intel briefing
Core summary
One sentence to understand this update
Ollama's v0.31.2-rc2 release now enables iGPU multimodal projector offloading with fit padding, resolving previous limitations on non-Metal iGPUs.
Impact & opportunity
What this could mean
This update improves performance and broader compatibility for multimodal models on integrated GPUs, particularly for users running Ollama on non-Metal systems.
Source
View original