問題文
A team reuses one intermediate frame in five later queries. The frame is too big to fit entirely in memory, and the team does not want to compute it again each time. They choose the level that keeps what fits in memory and puts the rest on local disk. What does that avoid?
選択肢
- The level that keeps blocks in memory and spills the rest to disk avoids the network transfer that a memory only level would cause.
- The level that keeps blocks in memory and spills the rest to disk avoids the wasteful rework that a memory only level would cause.
- The level that keeps blocks in memory and spills the rest to disk avoids the silent truncation that a memory only level would cause.
- The level that keeps blocks in memory and spills the rest to disk avoids the immediate halt that a memory only level would cause.