The SoC having 400GB/s of memory bandwidth doesn't mean that any individual CPU core can saturate 400GB/s of memory bandwidth (or even if all the CPU cores combined can achieve that). The SoC's memory bus is also feeding the GPU, and GPUs tend to be _very_ memory bandwidth hungry (see other discreet GPUs pushing over 1TB/s of memory bandwidth).
The CPU performance, even where memory IO limited, is more likely limited by how well it can prefetch memory and how many in-flight reads & writes it can do. A straight memcpy benchmark might be part of this suite, but that'd also be basically the only workload where a CPU core (or multiple) could come anywhere close to hitting 400GB/s of bandwidth. Otherwise memory _latency_ will be the bigger memory IO limitation for the CPUs, and that likely hasn't drastically changed with the new M1 Pros & Max's (there may be some cache layout changes which would shift some of the numbers, but DRAM latency is likely unchanged)
A single firestorm (performance) m1 core in the m1 Mac mini can sustain ~60Gb/s read from ram. I have no doubt that 8 of them could come close to if not entirely saturate 400GB/s.
The CPU performance, even where memory IO limited, is more likely limited by how well it can prefetch memory and how many in-flight reads & writes it can do. A straight memcpy benchmark might be part of this suite, but that'd also be basically the only workload where a CPU core (or multiple) could come anywhere close to hitting 400GB/s of bandwidth. Otherwise memory _latency_ will be the bigger memory IO limitation for the CPUs, and that likely hasn't drastically changed with the new M1 Pros & Max's (there may be some cache layout changes which would shift some of the numbers, but DRAM latency is likely unchanged)
For an example of this in a different product: https://www.anandtech.com/show/15578/cloud-clash-amazon-grav...
A single graviton2 CPU can "only" achieve 18-36GB/s of memory bandwidth even though the package in total can hit 200GB/s