People who don't like the results always like to claim that... but it's awfully hard to find proof.
Though, part of it I think may just be Geekbench's design as a microbenchmark that doesn't do long-term strain or attempt real-world tasks.
Though, part of it I think may just be Geekbench's design as a microbenchmark that doesn't do long-term strain or attempt real-world tasks.
> Geekbench began as a benchmark for Mac OS X and Windows and was created by John Poole who ran the now-defunct Geek Patrol website, which reviewed hardware and software designed for Macs, and featured editorials and interviews of interest to the Mac community. [0]
Basically, imagine an Nvidia enthusiast writing a system-agnostic GPU benchmark. I'm not as militant about rejecting Geekbench as some others, but you can't pretend the correlation is hard to make.
Regardless though, it's not hard to see Apple being 1-2 years ahead of Samsung's output. If the iPhone is being manufactured on recent TSMC silicon, it will annihilate anything Samsung's fabs are capable of producing.
This is more about Qualcomm being uncompetitive performance-wise, which has been true pretty much since the start of the smartphone wars.
The S23 Ultra went Qualcomm-only for the first time. In the year running up, they've been slowly expanding Qualcomm to more markets instead of Exynos.
It seems like they're gradually phasing out Exynos, at least from the higher-end phones -- since they've always struggled to produce an SoC competitive with even Qualcomm, which is already behind the curve compared to Apple.