Apple's recently introduced M3 Max processor sets the performance bar high, sending a clear message to its competitors. This chip excels not only in CPU performance but also in GPU performance, posing a challenge to Nvidia and AMD's high-end graphics cards.
Subjected to the GFXBench Aztex test, the M3 Max's 40-core graphics processor achieved an average of 443 fps at a 1440p resolution, totaling 28,517 frames. For comparison, this score is just slightly below a desktop RTX 3080 graphics card. The RTX 3080, in the same test, achieves an average of 440 fps, totaling 28,318 frames. In fact, it comes quite close to the laptop version of the RTX 4080. The RTX 4080 scored an average of 463 fps, totaling 29,759 frames.
The scores above are the result of following GFXbench query.
GFXbench Aztec Ruins High Tier Offscreen Result
In the GFXBench test, there can be inconsistent results.
Apple's M3 Max GPU is only about 7% slower than the laptop version of the RTX 4080, based on these scores. However, it's important to note that there's an issue with inconsistent scores provided by GFXBench. In this test, the M3's 10-core GPU could only achieve a score about 7.4% higher than the M2, which has the same number of cores. It's worth mentioning that the GFXBench software has not been updated for over a year. Therefore, software optimizations are needed to fully demonstrate the new GPU's capabilities.
Another aspect to consider is that Nvidia's cards operate in DirectX, while the M3 Max runs on Apple's proprietary Metal API. Hence, the close scores could be attributed to API differences. It's important to highlight this because, in the past, even the M2 Max struggled to deliver playable frame rates in high-resolution settings for older games like Shadow of the Tomb Raider. At the end of the day, dedicated graphics cards are specifically designed for extra performance, while the M3 Max's GPU, despite being powerful, is integrated into the processor, so expecting the same level of performance wouldn't be accurate.