Ammonia (NH3) fuelled engines have emerged as a promising route toward net-zero emission targets due to NH3’s carbon-free nature, ease of storage, and established handling infrastructure. However, the low laminar burning speed and narrow flammability limits of NH3 pose a significant combustion challenge, which can be addressed through hydrogen (H2) co-fuelling. For practical implementation, on-board H2 production via thermal catalytic cracking of NH3 is an attractive solution, as it eliminates the need for external H2 storage and associated handling and capital costs. Previous studies by the present authors identified a lean operating strategy that achieves an equimolar ratio of NOx and unburned NH3 (α NH3NOx ≈ 1), enabling complete conversion to nitrogen and water vapour when coupled with a Selective Catalytic Reduction (SCR) system. This strategy was further validated using cracked NH3 derived H2 in place of bottled H2 through an on-board cracker, thereby representing a practical system configuration. However, the required H2 fraction, and consequently the size and power demand of the onboard cracking system, is strongly influenced by engine architecture and operating conditions. The present study investigates the effect of compression ratio (CR) and stroke length, on H2 fraction requirements to achieve an optimum α of unity in an externally boosted SI engine. Results demonstrate that the high CR = 17.5, long stroke configuration reduces H2 enrichment by 50–60% compared to a low CR = 12.5, short-stroke engine architecture, allowing smaller onboard H2 generation systems. At high-speed, high-load conditions, it achieves over 45% thermal efficiency with stable NH3 combustion and no H2 supplementation, maintaining an α ≈ 1. Across the full operating map, NOx emissions comply with IMO Tier III and EPA Tier 4 norms, demonstrating near-zero-emission operation.