An analog-to-digital-converter (ADC)-based bit-error rate (BER)-aware blind receiver is presented to maximize the BER performance under a given low ADC resolution. The adaptation engine simultaneously adjusts the equalizer coefficients to compensate for the channel dispersion and calibrates the reference levels of the ADC to maximize the BER performance of the receiver. BER-aware adaptation is achieved through selective updates using the steepest-descent least-mean-square (LMS) algorithm. A time-interleaved 2 ×-oversampling ADC followed by a 2 × 1 multiple-input multiple-output (MIMO) equalizer improves the BER performance without sampling phase dependency. Simulation results reveal that the proposed design achieves a 4 dB performance improvement at a BER of 1 ×10^-5 over conventional designs in a 27” backplane channel at a data rate of 28 Gb/s.