Yes, if you look at the last chart at anandtech.com/show/16226/apple-silicon-m1-a14-deep-dive/3 (which I'm not allowed to post a link to since I don't comment here much), Qualcomm has always gone for a higher performance to energy ratio, whereas Apple went for more performance and didn't care about battery as much. Given Qualcomm and ARM's past focus on battery as much as performance, I'm sure that will still be the case with the new 888.

These mobile chipsets are really powerful these days, the 865 gets a Geekbench score about the same as a mid-range core i7 like the 10510U, which takes much more battery at 15W. That's why I gave up on Intel years ago, only using ARM for the last five years, including for heavy developer workloads like building software that takes hours to compile.

Apple has shown with the M1 that ARM is the winner for mobile computing, including laptops and work tablets, and we'll soon see what their desktop chip in the ARM iMac does. If I were in the stock market, I'd be shorting Intel heavily at this point.

Coming out swinging very poorly describes what is happening here. Qualcomm had a delicate balance to manage this time round. The Snapdragon brand has become associated with solid graphics performance and it was natural for Qualcomm to want to build on that reputation.

What was on offer from ARM with the X1 core, however, was a very large peak performance increase. Interestingly, Qualcomm has decided not to sqeeze all the performance available out of the SD888. In fact, despite the powerful X1 core it looks like the CPU cluster of the SD888 will have only modest power requirements in line with the existing SD865. Where Qualcomm is willing to spend a bit more energy, although not a lot, is on the new Adreno 660 GPU (and Video and Display control unit) that has received a major upgrade.

One problem that Qualcomm has had to deal with is an unremarkable 5nm silicon process node from Samsung. A better node would have allowed for additional power savings permitting either a flagship SoC with lower power requirements in 2021 than in 2020 or a SoC offering a 35% increase in single threaded peak integer performance rather than a measly 25%! But those possibilities weren't on offer with the Samsung 5nm LPE node.

The bottom line is Qualcomm has done the smart thing, given the constraints, and offered users as much performance as it could without overly elevating power requirements. That represents an exercise of great self-restraint. Qualcomm isn't waving its fists around but rather dancing like a butterfly and reserving its sting for when it really matters.