This paper proposes a novel scalable Multiple-Input Multiple-Output (MIMO) detector that does not require preprocessing to achieve good bit error rate (BER) performance. MIMO processing is a key technology in broadband wireless technologies such as 3G LTE, WiMAX, and 802.11n. Existing detectors such as Flexsphere use preprocessing before MIMO detection to improve performance. Instead of costly preprocessing, the proposed detector schedules multiple search passes, where each search pass detects the transmit stream with a different permuted detection order. By changing the number of parallel search passes, we show that this scalable detector can achieve comparable performance to Flexsphere with reduced resource requirement, or can eliminate LLR clipping and achieve BER performance within 0.25 dB of exhaustive search with increased resource requirement.