Benchmark: KV-cache save/restore in llama.cpp on Apple Silicon — TTFT, break-even, portability matrix, energy. Full data + harness.

apple-silicon benchmark kv-cache llama-cpp llm-inference
1 Open Issue Need Help Last updated: Jul 6, 2026

Open Issues Need Help

View All on GitHub
help wanted

Benchmark: KV-cache save/restore in llama.cpp on Apple Silicon — TTFT, break-even, portability matrix, energy. Full data + harness.

Python
#apple-silicon#benchmark#kv-cache#llama-cpp#llm-inference