Should You Split Decode? Disaggregation Across Future Models and Accelerators

arXiv:2611.08299v1 Announce Type: new Abstract: For large language model (LLM) inference, prefill/decode disaggregation (P/D) is deployed in production to isolate phases with different resource demands. Attention/feed-forward network disaggregation (AFD) splits decode again, but adds communication, placement constraints, and another provisioned pool. We ask whether role-specific placement repays this cost as models and accelerator roadmaps change. Infrastructure is committed before future models and hardware can be benchmarked. We present SiliconForge, a calibrated model that varies the serving decomposition and hardware assignment independently. We compare P/D and P/D+AFD across four model directions, four accelerator classes anchored to current and announced parts, and two fabrics.