Good stuff
I approve.
Benchmark software operation Futuremark has said it believes Samsung and HTC have attempted to rig the results its graphics testing app, 3DMark Android, yields when run on four of the two manufacturers’ most popular devices. According to Futuremark, “when a device is suspected of breaking our rules it is delisted”. Among those …
The fraction of people who consult benchmarks before buying is fairly small, and tend to read the likes of Anandtech. Of those people, I suspect a majority would have heard about these benchmark shenanigans, and would have quickly found online resources that give alternate measures - or a qualitative assessment - of a phone's performance.
So it would it seem that all Samsung and HTC's goosing of the benchmarks can only mislead a small sliver of their potential market.
and manufacturers rely on abovementioned people to bleed a steady stream of cash (buy, buy, buy). So the game continues. I haven't heard any apologies from the big boys how awfully sorry they are for this inadvertent design error. I guess they decided so long the story doesn't go mainstream, it's best just to pretend it never happened. And it works.
Moving these phones to the bottom of the list makes the list somewhat useless as I'm pretty certain these phones would ride quite high anyway. Wouldn't it be better if the benchmark was changed to penalise shorter battery life when achieving these higher speeds? Personally, I'd favour 20% extra battery life over 20% extra speed and I'd like to see benchmarks and tests that reflect this.
While Samsung have been caught bumping GPU clock speed, it's not the only way to game the scores. Nvidia used to get caught regularly dropping in custom shaders, dropping precision and changing texture modes in benchmarks. Sometimes that had zero effect on render output and is actually an acceptable approach in non-benchmarking contexts, drivers usually have game specific hacks built in - nowadays with en\disable control.
Even GPU overclocking isn't always easy to detect, or likely to decrease battery life enough for certainty.
It's a repeat of the NVidia+ATI+others situation, with the same solution - repeated public exposure till they stop.
Anybody that buys a PHONE because it is 2fps faster than another, needs to get a life! A phone is for communications, not for playing games. If you want to play games on the move, buy a DS or PSP! You might find that you actually get decent battery life out of the phone then!!!! As for those that play games on a tablet that uses motion to control the "character/vehicle" in public such as on a train etc, need to realise that you look a complete pillock sitting there tilting yourself from side to side!
I don't buy a PHONE, I buy a battery which has a flash light attached to it. It's a useful thing to carry with me in these dark November evenings.
That it also allows me to consult the internet and make telephone calls, tune my guitar and act as a calculator is a bonus. It's an alarm clock, a handy unit converter, an FM radio, and it makes a fair fist of being a camera, too.
I also own a real flash light (it takes 2x AA batteries and everything!), calculator (with graphs and Tippex decoration!), camera (the lens moves in and out!) and laptop (seems to think its a fan heater! Plays DVDs!)... but sadly I've lost my rucksack, so I have to leave them at home.
Hmm, this can't be right: My alarm clock thinks it's an FM radio. Freaky.