Samsung Denies Rigging Galaxy S4 Benchmarks
Synthetic benchmarks are not the ultimate judge of a device's performance. It's much more about real-world performance, which may be a key point that Samsung has forgotten when it comes to the Galaxy S4. Samsung has been accused of boosting the performance of its newest flagship handset by allowing its hardware to run faster in specific benchmark tests than it would in everyday use; a fact the electronics maker denies.
AnandTech began investigating the matter earlier this week, prompting forum goers on the website Beyond3D to find that the GPU in certain Galaxy S4 models ran at higher clock speeds during benchmark tests. These clock speeds weren’t attainable outside these benchmark apps, according to the reports.
MORE: Best Smartphones 2013
Specifically, the GPU issue applies to the international Galaxy S4 model which runs on Samsung’s own Exynos 5 octa-core processor. When operating in synthetic benchmark apps such as AnTuTu, Quadrant and GLBenchmark 2.5.1, the GPU ran at 532MHz, although it typically reaches 480MHz when performing everyday tasks. According to AnandTech, this is a boost of nearly 11 percent.
As for the CPU performance in both the international and Qualcomm-powered U.S. model, the report says that these benchmark apps cause the processor to jump to the highest clock speed possible—regardless of the workload.
Samsung, however, denies that it had altered the Galaxy S4’s performance in any way.
“[We] did not use a specific tool on purpose to achieve higher benchmark scores,” the company said on its Korean Web page.
The Korea-based manufacturer continued to say that the S4 operates at 533MHz “at its best performance,” and that certain full screen apps such as the camera, browser, video player and benchmarking tools require the highest performance available.
Meanwhile, AnandTech reportedly found strings of code that indicated specific benchmark tests were configured to perform at a faster speed, although Samsung has yet to directly respond to that allegation.
If these accusations prove true, Samsung’s alterations won’t necessarily affect the Galaxy S4 during everyday use. However, it does mean that standard benchmark results cannot be trusted as an objective means of deciding how the Galaxy S4 compares to its competitors. And it could mean some reviews of the smartphone may be unfairly skewed toward the Galaxy S4.
While we do use synthetic benchmarking on our reviews, we have also developed unique tests that more accurately reflect real-world tests, such as the LAPTOP Battery Test.