So, check out this useful “Compare Mac Models” page (h/t @gruber). Scroll down to the “Compute” and compare the stats. Go on, do it, I’ll wait. See anything that makes you go WTF?!? I did.
3/ The kind of heavy-duty work these things are designed for (LLMs, big-video, etc) all tend to work with huge blobs of data, so in many cases the memory bandwidth is the limiting factor. Old programmer mantra: “All computers wait for data at the same speed.”
@timbray the M6 is inferior in almost everything, but it's also half the price of the M5 Max...
I think the 6 stands for "2026"
@johnefrancis Huh… I'd have thought the performance increased as the number after the “M” did.
@timbray Apple’s ”M” chips come in multiple sizes for (most) generations. (Replace n with 1-6.)
* Mn - The plain, no-suffix M chip. This is the smallest and cheapest, with the narrowest busses and the least cores, etc.
* Mn Pro - The next size up: more cores, wider busses, more PCIe lanes, etc.
* Mn Max - The biggest “single chip” M chip (though the M5 Max does use separate CPU and GPU dies now).
* Mn Ultra - Two Mn Max chips stuck together, so double everything the Max has.
2/ It’s the Memory Bandwidth numbers, “Up to 170 GB/s” for the M6, “Up to 614 GB/sec” for the M5 Max/Pro models. This is a big deal? Computers pull data from memory into the CPU, crunch it, write it back out. But they can crunch many times faster than the memory can deliver it. So the effective performance depends a *lot* on memory bandwidth.