I've heard that explanation but there in lies my point. We all know that real world performance don't equate to the theoretical numbers. But in terms of what is typically marketed, game companies never get in the weeds of that kind of detail. Every computing system has bottlenecks that prevent it from achieving the theoretical specs but that never stops companies from presenting the theory/dream. Who remembers that the PS2 was
marketed as a 6.2GFLOP machine which was an astronomically high number for the day. Did real world games ever get close to that? No. How about the fact that in its initial presentation to the public, the PS3 (not PS4) was marketed as a 2 TFLOP machine! Yeah right! So what I'm saying is that figure of +45% rendering performance doesn't make sense from a marketing perspective when:
- The figure itself is pretty low, especially in the context of what the PS4 Pro was relative to the PS4. From a marketing perspective, it's not a great metric to even present in the context of overall GPU perf
- The typical metric used for marketing GPU power has been TFLOPs and by their own data, the TFLOP figures point to a higher relative performance (+65%)
Again, the term "rendering performance" has NEVER been used as a performance metric with the arrival of a new console. Rendering is just one aspect of the graphics pipeline and don't tell the whole story of what the overall graphics performance will be. It just doesn't add up.