It doesn’t look like we mentioned this one yesterday, so I’ll start from the beginning. Essentially, it was alleged that Samsung cheekily allows the Galaxy S4’s GPU and CPU to run at higher clock speeds in specific benchmarking apps, yielding better results than users might expect in real life and possibly skewing reviews.
Samsung, which – to its credit – has a history of quickly responding to potentially negative press in blogs (I’m thinking of Exynos-related security concerns, in particular), has claimed innocence.
A forum post led to a remarkably in-depth investigation by AnandTech, which concluded that the Exynos-fuelled Samsung Galaxy S4 GPU is able to run at 533MHz in certain benchmarking tools, though it’s normally locked to 480MHz.
It’s a similar story, seemingly, in CPU performance measuring for both the Exynos and Qualcomm-based Samsung Galaxy S4s in the US.
A Sammo statement (translated from Korean-language Samsung Tomorrow) reads: “[We] did not use a specific tool on purpose to achieve higher benchmark scores. Under normal conditions, the Galaxy S4 operates up to 533MHz at its best performance."
Interesting stuff, but does it matter much? I can’t help but feel that this is largely the domain of super-nerds. I doubt we’ll see a mass return of Samsung Galaxy S4s on account of benchmark rigging.
Do you have a Samsung Galaxy S4? Are you troubled by the allegations, or thoroughly unmoved?