Allan Jones wrote:
Benchmark numbers in MacTracker show only a couple of percent difference between the fastest current iPad Pro, Mac Mini M4 Pro, and M4 Max Macbook Pro:
…
In fact, the scores are so close that it makes me question how effective GeekBench can be for testing Apple Silicon processors.
Those are single-core CPU benchmark results. The results there will be very similar between all of the levels of processors, because for a particular Apple Silicon generation, plain, Pro, and Max processors all share the same basic performance CPU core, and efficiency CPU core, designs.
If you are shopping for a Mac for general-purpose use, this is good news. Given enough RAM, a Mac that has a plain M4 chip may – for a lot of everyday productivity and light gaming uses – have performance that is virtually indistinguishable from that of a Mac with a M4 Pro or M4 Max chip.
What the higher-end chips bring to the table is more of various types of hardware units: CPU cores, GPU cores, display generators, etc. So you'll see multi-core CPU scores jump when you go from a plain M4 to a M4 Pro or Max – something that may be of interest to people who run long batch jobs that can make good use of all cores. Much general-purpose use depends more on single-core performance, or stresses the CPU so little that even a plain M-series chip will complete jobs "instantaneously", given sufficient RAM.
You also get stronger GPUs with the higher-level chips. But the Geekbench CPU scores don't reflect that.