On Neoverse-V1/V2, the SVE `CPY (immediate, merging)` instruction performs better than the SVE `CPY (immediate, zeroing) instruction. Optimizing `CPY(immediate, zeroing)` as `MOVI + CPY(immediate, merging) gets performance uplift of **12%** to **100%** in specific Java Vector API micro-benchmarks depending on the specific operation and data types involved.

Currently the SVE `CPY (immediate, zeroing) instruction is used in code generated by `VectorStoreMaskNode` and `VectorReinterpretNode`. Doing this optimization benefits all Vector APIs that generates these two IRs, such as `VectorMask.intoArray()` and `Vector.toLong()`.