Even if massive multiple-input multiple-output (MIMO) can theoretically bring huge benefits, it incurs substantial hardware complexity and expensive hardware costs. To address these issues while maintaining the system performance simultaneously, we develop an asymmetrical transceiver architecture for massive MIMO systems in this paper by releasing the shackles on radio frequency (RF) chains. Specifically, we first develop the architecture for the asymmetrical transceiver where the number of receive RF chains is different from that of the transmit RF chains. Due to this unequal number of RF chains, channel inconsistency appears. To overcome the channel inconsistency and thus fully harness the system's downlink data transmission capability, we propose two uplink-to-downlink channel transfer algorithms. The cost and power consumption models for the asymmetrical transceiver are also developed and the uplink signal-to-noise loss due to the channel inconsistency is investigated. Through analyzing system spectral, cost, and energy efficiency, we demonstrate that the proposed asymmetrical transceiver-based massive MIMO system can achieve excellent downlink spectral efficiency while maintaining a reasonable energy efficiency.