This seems to be the only way to combat ARM. I wonder how long amd takes to come out with their version of big.little.
I think AMD has handily circumvented this by having an x86 design that isn't a patchwork of Skylake improvements and 1956710151 attempted tweaks.
This whole Alder Lake story... looks impressively unimpressive so far. End of '21? Right.
Meanwhile we still haven't moved an inch from the reality of two years ago: Intel still can't push the performance CPU button on 10nm like they could on 14nm, they still have a vast range of new CPUs that have heavily increased TDPs and require highly aggressive burst modes to get their peak clocks for only an instant.
I've lost count of the amount of announcements on fantastic new architectures from Intel. Even Xe however has been nothing but powerpoint slides and some rebranded IGPs.
Alder Lake and Big Little on the performance segment sounds a lot like 'we want to extract the last few % from our big cores so we need a new number trick to fool the TDP budget'. Its a trick. Nothing more. There is nothing here that will give more net performance per watt, only net performance per watt per timeframe - they gain some minor versatility to push more power to a big core. Speak of making a mountain out of a mole hill.
This is not the way forward. This is Intel desperately clinging to old ideas because they're too deep in R&D hell.
www.tomshardware.com
