Search papers, labs, and topics across Lattice.
The paper introduces Fovea, a novel methodology for wafer-scale design-space exploration (DSE) that incorporates physical implications into the architecture selection process. By leveraging a workload-specific approach and Decision-Domain-guided cross-fidelity refinement, Fovea effectively navigates the complexities of coupled physical implications and reduces the computational burden of detailed evaluations. The results demonstrate that Fovea achieves up to 7.80x speedup while accurately recovering the designated-reference optimum across multiple design spaces and workloads.
Fovea achieves a remarkable 7.80x speedup in wafer architecture selection while ensuring optimal performance across diverse workloads.
Modern pre-silicon design-space exploration (DSE) follows a coarse-to-fine workflow: low-cost evaluators screen candidate spaces, while detailed evaluation is reserved for a shortlist. Wafer-scale systems strain both stages. Architecture choices induce coupled physical implications for reticle compliance, wafer tiling, die area, D2D capability, boundary access, and placement, so the design space cannot be treated as an unconstrained Cartesian product. Meanwhile, detailed evaluation is too expensive to cover the resulting space, whereas analytical-to-reference ranking inversions make a fixed shortlist unreliable. We present Fovea, a reusable methodology for workload-specific wafer architecture selection rather than a fixed wafer template. Fovea first performs physical-implication-aware design-space formulation to construct a distinct modeled-feasible space while preserving cross-dimensional trade-offs and applying only evaluator-preserving local reductions. It then performs Decision-Domain-guided cross-fidelity refinement. Paired in-domain calibration estimates workload- and space-specific analytical-to-reference disagreement, which parameterizes reference-consistent performance intervals and induces a Decision Domain for selective designated-reference evaluation. Under a valid domain-wide disagreement bound, this domain contains the designated-reference optimum; the sampling-based implementation is evaluated empirically on exhaustive-reference design spaces. Across ten reference-verifiable design spaces and seven LLM-training workloads, Fovea with 10% paired calibration recovers the exhaustive designated-reference optimum in all 70 evaluated pairs while achieving 4.13x average and 7.80x maximum end-to-end speedup.
