In high-concurrency LLM serving, model weights are rarely the only memory problem. KV Cache grows with requests and tokens, so PagedAttention and Continuous Batching are the real keys to vLLM throughput.
高并发大模型服务真正容易爆掉的不是模型参数,而是随请求增长的 KV Cache。理解 PagedAttention 和 Continuous Batching,才能看懂 vLLM 为什么能把吞吐做上去。