Parallel Merge

题面

并行合并两个已排序数组 A(M) 与 B(N)（float32，非降序）到输出 C(M+N)，强调并行策略（分治/对角线划分等）以避免串行依赖。

Implementation Requirements

• Use only GPU native features (external libraries are not permitted)
• The solve function signature must remain unchanged
• 合并结果写入 C

Examples

A=[1,3,5,7], B=[2,4,6,8] → C=[1,2,3,4,5,6,7,8]
A=[-1,1,3], B=[2] → C=[-1,1,2,3]

Constraints

• 1 ≤ M,N ≤ 50,000,000；M+N ≤ 50,000,000
• 输入已非降序；元素为 float32
• Performance: M=N=25,000,000