Multi-input multi-output (MIMO) antenna systems provide the user with additional degrees of freedom over traditional single antenna (SISO) system to enable optimum transmission. This paper focuses on quantifying the maximum energy efficiency of ldquomode-richrdquo MIMO wireless communication systems. We consider a MIMO-OFDM (orthogonal frequency division multiplexing) based software defined radio (SDR) where the user is allowed to dynamically change various system parameters such as bandwidth, transmit power, constellation size, channel coding rate, decoding algorithm, and MIMO antenna configurations. Based on these parameters, standard geometric programming problems are formulated to optimize energy efficiency of the system given quality of service (QoS) constraints. The result is a comprehensive MIMO link adaptation strategy that can optimally match the communication mode to the user QoS requirement and the channel condition. Results show that an optimized MIMO system can provide an order of magnitude improvement in energy efficiency over a static strategy.