When running scimark on aarch64:

 ;; B16: # B17 <- B21 top-of-loop Freq: 2305.21

  0x000003ffa126f710: add w17, w11, w12 ;*iadd {reexecute=0 rethrow=0 return_oop=0}
                                                ; - jnt.scimark2.FFT::transform_internal@243 (line 129)

  0x000003ffa126f714: nop
  0x000003ffa126f718: nop
  0x000003ffa126f71c: nop ;*iconst_2 {reexecute=0 rethrow=0 return_oop=0}
                                                ; - jnt.scimark2.FFT::transform_internal@238 (line 129)

 ;; B17: # B32 B18 <- B25 B16 Loop: B17-B16 inner Freq: 3056.06

  0x000003ffa126f720: lsl w16, w17, #1 ;*imul {reexecute=0 rethrow=0 return_oop=0}
                                                ; - jnt.scimark2.FFT::transform_internal@244 (line 129)

The 3 nops are added by the code that align loops: the top of loop block is first encountered and its alignment is set, the loop head is later encountered through the backbranch of an outer loop.