llama.cpp update cuts VRAM usage by 1.2 GB via smarter logits allocation
The llama.cpp PR #23861 changes the context to reserve logits space only for active sequences, trimming unnecessary buffer allocations. Combined with the -ub 2048 setting and Multi‑Threaded Parallel (MTP) decoding, the change frees roughly 1.2 GB of VRAM, extending the benefits of recent KV‑cache quantization for multi‑GPU deployments.