M5 Pro and Max unveiled

[dada_dave]

Indexing can take up to a week. Also are you adding high power mode to this chart? Because that dramatically increases power consumption usually, and usually it doesn't actually add performance but rather sustains it.

I believe NBC used high power mode for this test, which I believe is their standard and a lot of what NBC tests is sustained performance (not everything admittedly and I don't chart the results from the really long ones). However, even high power mode shouldn't cause the M5 Pro to draw 100W in CBR24 while getting that low a score. I can compare that result to other Apple laptops in the NBC data set, also run at high power, and that's just not reasonable.

It's indexing most likely. There are fluctuations in early GB tests.

It's certainly possible, why I brought it up as well, but, unless Apple screwed up, the review units sent out should not have been indexing (they are supposed to be set up so the reviewers can start right away as usually they have very limited amount of time with these machines before embargoes are up) and while NBC didn't see that behavior on the M5 Pro model oddly, they did see exactly that in the M5 Max 14" model:

The MacBook Pro offers three different performance modes: Automatic, High Power & Low Power and we have summarized the effects of the three different modes in the table below. All three modes are also available on battery power with the same performance figures. If you want the maximum CPU and GPU performance, you must use the High Power mode, so we used it for our tests as well. However, the maximum power limits shown are only reached for one or two seconds on the 14-inch model and it will almost immediately throttle down. While the GPU performance was reproducible, we had issues with the CPU performance and encountered Cinebench 2024 Multi scores ranging between ~1400 and the maximum of 2073 points with the consumption dropping below 40 Watts at times, even though the testing conditions for the runs were pretty much identical. We are not really sure what is going on here, maybe Apple will fix this behavior with a software update. As of now, you will not get very consistent CPU performance under sustained workloads.

emphasis mine. That's exactly what @exoticspice1 encountered. In the NBC Max review, I originally took that to be throttling and maybe it is! Maybe it's coincidence since after all @exoticspice1 is getting high variability in the machine NBC claims they didn't get it in (although at least one of NBC's measurements on that machine is the one I'm most questioning!). It's all very odd.

Also GB seems to fluctuate like crazy between different machines regardless of indexing. Especially the subtests. When @leman and I did violin plots of the subtests, the outliers were horrible both in number and extent - those charts are floating around these forums somewhere. One expects some differences between machines, silicon lottery and all that, but GB is very sensitive to it. The median is fine, but I suspect the very short runtimes is the problem there. But yes, some of this early stuff is going to be from people benchmarking a machine that is still indexing. I still use GB, but yeah ... I'd prefer slightly longer tests to cut down the noise, but maybe their general audience wouldn't.