Benchmarks of the newly unveiled Apple M4 Max chip have started making waves online. Based on leaked Geekbench scores, the M4 Max single-handedly outperforms the latest processors from Intel and AMD—all while using noticeably less power.
Notably, the Apple M4 Max, benchmarked on the new 16-inch MacBook Pro, scores 4,060 points in the single-core test and 26,675 points in the multi-core test. In comparison, last year’s M3 Max scored 3,070 in single-core and 21,405 in multi-core tests, according to the Geekbench database.
This means the M4 Max offers about a 32% boost in single-core performance and roughly a 25% increase in multi-core performance over the M3 Max.
Not only does Apple’s new chip outperforms its predecessor by a solid margin, but it also edges out Intel’s Core Ultra 9 285K and AMD’s Ryzen 9 9950X processors. Here’s how the benchmarks compare:
Processor | Geekbench Single-Core | Geekbench Multi-Core |
---|---|---|
M4 Max | 4,060 | 26,675 |
Ryzen 9 9950X | 3,359 | 20,550 |
Intel Core Ultra 9 285K | 3,449 | 23,024 |
M3 Max (16-core) | 3,070 | 21,405 |
Based on benchmarks, the M4 Max outperforms the Ryzen 9 9950X by 20.8% in single-core and 29.8% in multi-core tests. It also scores 17.7% higher in single-core and 15.9% higher in multi-core than the Intel Core Ultra 9 285K.
Even when AMD’s Ryzen 9 9950X ‘Zen 5’ is overclocked to 6.0GHz, achieving a multi-core score of 26,047, the M4 Max still holds the single-core performance crown.
In terms of specifications, the M4 Max features a 16-core CPU with 12 performance cores and 4 efficiency cores. Built with advanced 3nm technology, the Apple chip is also highly power-efficient—a key factor behind the new MacBooks’ claimed 24-hour battery life, the longest ever on a Mac.
In short, the M4 Max sets a new standard for those pushing their workflow limits, making it ideal for complex tasks like video editing, 3D animation, and graphic design. Apple has also doubled down on AI, with a Neural Engine twice as powerful as the previous generation and upgraded machine learning (ML) accelerators in the CPU.