With　the　advancement　of　technology,　the　energy　consumption　of　computing　devices　continues　to　rise.　As　a　new　paradigm　to　address　energy　challenges　in　error-tolerant　environments,　approximate　computing　holds　significant　importance.　This　paper　proposes　an　approximate　multiplier-the　Low-Carry　Unsigned　Approximate　Multiplier　(UAM8-FP),　an　8×8　approximate　multiplier　design.　The　key　innovation　of　UAM8-FP　lies　in　its　exclusive　use　of　carry-free　compressors,　which　drastically　reduces　carry　propagation　during　partial　product　reduction　and　significantly　accelerates　partial　product　accumulation.　The　second　core　feature　of　UAM8-FP＇s　compressor　strategy　is　its　phase-dependent　adaptive　compression　technique,　where　compressors　of　varying　approximation　densities　are　deployed　based　on　the　sparsity　of　＇1＇s　at　different　reduction　stages.　This　approach　not　only　enhances　computational　speed　but　also　minimizes　precision　deviation　probability.　To　further　optimize　performance,　a　third　strategy　is　introduced:　customized　compressor　allocation,　where　exact　and　approximate　compressors　are　dynamically　selected　based　on　partial　product　distribution.　By　integrating　carry　reduction,　staged　approximation,　and　column-wise　optimization,　UAM8-FP　achieves　a　balanced　trade-off　among　speed,　hardware　cost,　and　accuracy.　Compared　to　existing　approximate　multipliers,　it　demonstrates　strong　competitiveness,　offering　both　academic　and　practical　value.　©　2025　IEEE.