Intel’s VP Robert Hallock claims that better CPU-level optimization could boost game performance by about 30%, a point supported by experts who note that current games often underutilize Intel’s architecture features like efficiency cores and cache optimizations. While AMD’s Zen architecture and X3D cache technology currently provide advantages due to broader software optimization, future Intel products with improved cache designs may help close the performance gap if developers optimize games more effectively for Intel CPUs.
Intel’s VP Robert Hallock recently claimed that games could run about 30% faster if they were better optimized for PC, sparking debate about whether this is a genuine observation or an excuse for Arrow Lake’s underwhelming performance compared to AMD’s X3D technology. The discussion highlights Intel’s efforts to optimize games at the CPU level, suggesting that current games may not fully leverage specific CPU architectures, much like how GPU drivers are optimized for gaming.
Alex agrees with Hallock’s statement to an extent, pointing out that better optimization, such as writing cache-friendly code, could indeed yield significant performance improvements. He notes that in the past, instruction sets were more critical, but modern games often avoid using the latest CPU instruction sets for compatibility reasons or development convenience. This lack of optimization results in missed opportunities for performance gains on CPUs like Intel’s.
Another key point Alex raises is the underutilization of Intel’s efficiency cores (e-cores) in gaming. He believes that explicitly targeting these cores for background tasks like decompression or shader compilation could offer tangible benefits, but current game engines typically do not exploit them. This contrasts with AMD’s approach and the widespread adoption of Zen architecture in consoles and PCs, which tends to shape game development priorities.
Oliver adds that optimizing games for different CPU architectures takes time and resources, and since many games are designed with Zen in mind, they naturally perform better on AMD hardware. He views AMD’s X3D cache technology as a brute-force but effective solution to cache misses, suggesting Intel might need to develop a comparable or transformative technology to remain competitive in benchmarks and real-world performance.
Both agree there is merit to Hallock’s claim about optimization driving performance improvements, especially regarding cache management. They speculate that future Intel products like Nova Lake, rumored to feature larger cache tiles, could address these issues more directly. However, the consensus is that current performance gaps are partly due to software not being fully optimized for Intel’s architectures, and better optimization could unlock substantial gains.
