IT'S AN OLD STORY. Get your new handset, take it home, marvel at its immense speed compared to your last one, add a few apps, and watch it grind to a halt.
"But why?" you wail to yourself, "This is the fastest gadget on the market."
The truth is out there. Most phone benchmarks are, shall we say, 'adjusted' at an OEM level. An in-depth report from Anandtech has shown that manufacturers are using all kinds of trickery to get the best result.
The research concentrated on two popular chips used in most of the major smartphones now. In the case of the Exynos 5410, as is used in the Samsung Galaxy Note 10.1 tablet, the chip has been set to raise thermal limits if it detects a benchmarking exercise.
Meanwhile on the Samsung Galaxy S4, both chips - the Qualcomm Snapdragon 600 and Exynos 5410 - have been set to max out CPU frequency and throughput as soon as a benchmark is detected, with the Snapdragon set to trigger all four cores at once, which would never happen under normal conditions.
But it doesn't stop with Samsung. In fact only Apple and Motorola haven't partaken in this exercise in virtual anabolic steroids. Nexus units also use 'real' benchmarks and Anandtech speculated that it could be the honesty that have made the Nexus phones disappointing for some.
The reality is, however, that the performance boosts afforded by these modifications are relatively tiny, which serves to show just how cutthroat the benchmark game has become. Could we be entering an era where dope testing becomes part of the quality control process? And if Apple can produce silky smooth performance without massaging its figures, why can't everyone else? µ