MLPerf Training v6.0 introduces GPT-OSS 20B, a new sparse Mixture-of-Experts (MoE) pretraining benchmark designed for accessibility on single 8-GPU nodes.