Samsung - where performance matters
#1
This is pretty crazy (sad). Samsung apparently built code into their SoCs (System on a Chip) in phones and tablets to detect when benchmark software is being run (specific benchmark app names are embedded in the firmware), and bump up the clock speed of the CPU when it is. In other words...they get high benchmark scores that will NEVER be used for apps or games. Why? Because running at such a high clock speed would lower battery life. Talk about gaming the system...

http://www.anandtech.com/show/7187/looki...galaxy-s-4
Ex SWG, L2, CoH, Wow, and War
Currently PvPing in the stock market
Reply
#2
That's lame... I still <3 my S4 though.
I don't own kid gloves.

Steam Friend Code : 1636490
Reply
#3
Obviously Samsung denies.

Link

Quote:In response, Samsung said that "under ordinary conditions," the Galaxy S 4 can attain maximum GPU frequency of 533MHz.
"However, the maximum GPU frequency is lowered to 480MHz for certain gaming apps that may cause an overload, when they are used for a prolonged period of time in full-screen mode," Samsung said. "Meanwhile, a maximum GPU frequency of 533MHz is applicable for running apps that are usually used in full-screen mode, such as the S Browser, Gallery, Camera, Video Player, and certain benchmarking apps, which also demand substantial performance."

Either way doesn't really matter to me... the S4 has been great for me so far and is a great phone IMO.
I don't own kid gloves.

Steam Friend Code : 1636490
Reply
#4
It now looks like all Android smartphone vendors do this, with the exception of Google/Motorola. Certainly all the South Korean/Chinese ones. Pretty sad.

http://www.anandtech.com/show/7384/state...benchmarks
Ex SWG, L2, CoH, Wow, and War
Currently PvPing in the stock market
Reply
#5
Especially as benchmarking numbers matter to approximately .01% of smartphone consumers.

Also, whoever understands everything in that article needs to go out and get laid.
Reply
#6
I don't see the problem.

The CPU can operate in 2 modes.
They whitelist particular apps that they know exist and that they know need higher power to use the higher power mode. All others get the lower power mode.
Benchmarking programs want to measure maximum possible speed, so they get whitelisted.


The real question is whether there's an API for this whitelist and if you can add your own app to it if need be.
Reply
#7
How is it not a problem? They are gaming the system. Basically they are deliberately fooling certain benchmark apps to believe they can achieve a certain performance, when in fact no regular app can get that performance.

Or to put it another way. A Samsung S4 says it can achieve 60 FPS on a graphics-intensive benchmark. A Nexus 4 says it can achieve 50 FPS on the same benchmark. Samsung says, "the S4 has the best performance of any Android phone, buy us!". And a certain segment of the population does. Only...the S4 would actually score the same, or maybe lower, as the Nexus 4 without cheating. You don't see a problem with that? How about if this was a gaming PC?

I could personally care less, since I don't buy based on benchmarks, and even if I did, the iPhone 5S (and even the year old iPhone 5) crushes the latest Android smartphones even with their faked scores... Wink

It's just another example to me what a sad, unscrupulous excuse for a company Samsung is (and apparently other Asian companies too). This is on top of paying people to go online and praise their products and trash others, not to mention the whole copying thing.
Ex SWG, L2, CoH, Wow, and War
Currently PvPing in the stock market
Reply
#8
Grieve Wrote:This is on top of paying people to go online and praise their products and trash others

I know, if only they had spent their time founding a fanatical religion they would have saved so much money!

Grieve Wrote:Or to put it another way. A Samsung S4 says it can achieve 60 FPS on a graphics-intensive benchmark. A Nexus 4 says it can achieve 50 FPS on the same benchmark. Samsung says, "the S4 has the best performance of any Android phone, buy us!". And a certain segment of the population does. Only...the S4 would actually score the same, or maybe lower, as the Nexus 4 without cheating. You don't see a problem with that? How about if this was a gaming PC?

Actual non snarky question (i know in an apple-android thread??) about this, since I'm not a technophile. Movies are filmed around 30 FPS - why should we care about higher FPS - what is the difference when thats around what your brain processes? Whats the point of diminishing returns?
Reply
#9
Benchmark apps want to know what the maximum speed the hardware can handle is. They whitelist them so that they can see.

It's like complaining that using SpeedTest.net to measure your ping and speed to D.C. is cheating the system because really most of the time you're communicating with California. Yeah, well, Speedtest isn't generally used to measure "likely speed in most conditions". It's supposed to measure maximum speed.

Same for benchmark apps.


Maybe more to the point, I'll care more when I actually feel my CPU/GPU performance is a problem. Buying something that's faster than an S4 is a bit like going shopping for a video card and getting the latest top-of-the-line card. You'll spend more money getting something you don't really need.

(And really it's a lot sillier to do this for phones. You might keep your video card for 5 years and grow into it. 5 years from now you'll be on the iPhone 10s and your current iPhone will be lost in a drawer somewhere.)
Reply
#10
Jakensama Wrote:Actual non snarky question (i know in an apple-android thread??) about this, since I'm not a technophile. Movies are filmed around 30 FPS - why should we care about higher FPS - what is the difference when thats around what your brain processes? Whats the point of diminishing returns?
That would probably start a fistfight in PC gaming forums where this exact topic has always come up.

But I've never been able to see the difference in anything beyond about 30 fps so I've never understood it when people bragged that their system could do 300 fps.

Like < 10 is noticeably shitty. ~20 is just enough to make you think the animation isn't as smooth as it should be. 30+ all looks the same to me.
Reply
#11
Slamz Wrote:Benchmark apps want to know what the maximum speed the hardware can handle is. They whitelist them so that they can see.
That would only be valid if ALL phones whitelisted the apps. Then you'd have apples to apples (no pun intended!).
Benchmarks are, by implication, a way to compare one bit of hardware to another , whether it's a video card, a hard drive, a phone, or whatever. If not everyone is playing by the same rules, it's useless as a point of comparison.

And yes, 95% of consumers probably don't care, but quite a few people do, as shown by the fact there are are already 1000s of comments on the various stories on this around the web.
Ex SWG, L2, CoH, Wow, and War
Currently PvPing in the stock market
Reply
#12
Grieve Wrote:And yes, 95% of consumers probably don't care, but quite a few people do, as shown by the fact there are are already 1000s of comments on the various stories on this around the web.

56% of American adults own a smartphones. 1000s are, literally, an insignificant glitch.

Hundreds of thousands would be a minor outcry.
Reply
#13
Jakensama Wrote:Actual non snarky question (i know in an apple-android thread??)
Forget the thread topic, a non-snarky question from YOU is the real shocker! Wink

Jakensama Wrote:about this, since I'm not a technophile. Movies are filmed around 30 FPS - why should we care about higher FPS - what is the difference when thats around what your brain processes? Whats the point of diminishing returns?
I'm much more of a moviephile than a gamer, and 30 FPS is actually pretty crappy. We're just used to it. For the US, most movies are 24FPS, but there is a big move towards 48FPS now. Peter Jackson started it off with the Hobbit, which really freaked a lot of people out (because the hyper-realism made it look "un-movie-like"), but James Cameron is doing the next Avatar movies in 48FPS as well, so as more movies come out that way, it will get more accepted.

For games, it's probably more to do with the refresh rate of the monitor. 60Hz is supported by just about all modern monitors, but you can get 120Hz and 144Hz as well.

With PC games, it's more about balancing the frame rate against the resolution and graphics "stuff" you have turned on. In other words...if you only get 5FPS while all the fancy options turned on, you'd better off turning some stuff off (and maybe lowering resolution) so you can get up to 60FPS.
Ex SWG, L2, CoH, Wow, and War
Currently PvPing in the stock market
Reply
#14
What's the problem? Ill tell you this, side by side using applications that mater, the Samsung s4 blows away the apple iphone 5. This is based on in hand comparisons.
Maul, the Bashing Shamie

"If you want to change the world, be that change."
--Gandhi

[Image: maull2.gif]
Reply
#15
Correct me if I'm wrong but Samsung S4 = quad core.
Apple iPhone 5s = dual core.

That would explain why the Samsung S4 appears to be more powerful despite a lower clock speed. Android, at least, seems to be very multi-core / multi-thread friendly so most apps are probably benefiting nicely from the quad core processor.
Reply
#16
Yes, the S4 is quad core and has higher clock speed in most models (1.6 or 1.9). But the 5S is highly optimised and far, far more powerful than the S4. Even the year old 5 beats the S4 is most benchmarks. More cores and more clock speed does not equal better performance.

Maul, what apps are you comparing?
Ex SWG, L2, CoH, Wow, and War
Currently PvPing in the stock market
Reply
#17
Maps. GPS navigation, email, Facebook, good.
Maul, the Bashing Shamie

"If you want to change the world, be that change."
--Gandhi

[Image: maull2.gif]
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)