It?s becoming increasingly clear that there?s a gap between synthetic benchmarks and real world usage on mobile devices, a truth particularly evident on the Samsung Galaxy S4. Out of the box performance for the handset left a lot to be desired when it first shipped, although things did improve with subsequent over-the-air updates. Anandtech has discovered why there?s such a large discrepancy, though, with the site claiming that Samsung is artificially boosting benchmark performance.
Here?s the technical explanation: the GPU found in the Exynos 5 edition of the Galaxy S4 is typically clocked?at 480Mhz, even when running extremely demanding mobile games. But when someone runs a specific benchmark on the handset - AnTuTu or Quadrant, for example - the GPU suddenly clocks itself up to 532Mhz. That number doesn?t seem to be attainable in other applications either, and what makes the discovery even more damning is a ?BenchmarkBooster? string buried in code on the phone.
Samsung has already responded to the allegations, saying that the boosted GPU performance is available to any app that needs it ?under normal conditions?. That would include full screen apps such as the camera, browser, and video player, but the company didn?t address the specific code strings that seem to target benchmarking apps.
At the end of the day, how your phone performs in the real world is all that matters. Benchmarking tools can prove to be good comparison tools for different handsets, but do the raw numbers really matter if you?re seeing smooth performance in your day to day usage?
One thing's for sure -?If the allegations are?true, it?s an extremely?deceptive move of Samsung part, especially when its flagship devices typically rank near the top of performance tests.
Source: http://recombu.com/mobile/news/samsung-galaxy-s4-artificial-benchmarks_M19418.html
May the Fourth be with you James Righton finish line kentucky derby Iron Man 3 margaret thatcher MET GALA 2013
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.