This contribution reports the simulation results of applying SbTMVP to AMVP mode, which is studied in EE2 Test 2.2. Given a CU coded in AMVP with SbTMVP mode, the CU is predicted in a similar way as that of SbTMVP in merge mode except that the motion shift is signaled in the bitstream instead of being derived from left neighboring block. The motion shift comprises a derived MVP and a signaled MVD. The number of MVD is set to 16 if a picture whose POC distance between current picture and collocated picture is larger than 2. Otherwise, if the picture is a low delay picture, the number of MVD is set to 32. Otherwise, the number of MVD is set to 64. It is reported that on top of ECM-8.0, the overall coding performance impact for {Y, U, V, EncT, DecT} is {-0.07%, -0.02%, -0.04%, 104%, 100%} for RA and {-0.22%, -0.15%, -0.14%, 103%, 100%} for LB.