AWS Neuron 2.32 introduces expanded NKI programming, MXFP8 training kernels, and variable-size collectives for Trn2 and Trn

AWS Neuron 2.32.0 is now available, introducing NKI 0.6.0 with an on-device top-K instruction, a variable-length `all-gather `for kernels whose ranks hold differently sized tensors, and new loop constructs for data-dependent iteration. This release also introduces 13 new NKI Library kernels for Mixture of Experts (MoE) training and sparse attention, and a new Neuron Agentic Development skill that