AArch64: [vectorapi] Optimize Vector.reduceLane for SVE 64/128 vector size

XMLWordPrintable

    • aarch64
    • generic

      The performance of Vector.reduceLane(op=MUL/MIN/MAX/ADD), can be improved in 128-bit vector size by using low costs NEON instructions.

            Assignee:
            Eric Liu
            Reporter:
            Eric Liu
            Votes:
            0 Vote for this issue
            Watchers:
            2 Start watching this issue

              Created:
              Updated:
              Resolved: