Samsung boosts graphics in Exynos Galaxy S4 to manipulate benchmark scores
0. phoneArena 30 Jul 2013, 09:33 posted on
The Samsung Galaxy S4 is the best selling Android smartphone of all times and it’s scored tremendously well in benchmarks…
This is a discussion for a news. To read the whole news, click here
23. jan25 (Posts: 384; Member since: 26 Feb 2012)
i think we have gotten to the point where the little benchmark boost in score means nothing. it all now depends on the user experience and battery life.
25. dmat3889 (Posts: 14; Member since: 14 Mar 2013)
I get that the clock speed isnt true to real world speed but people can also OC thier phones to be at that amount.
Now whats the difference between this and say when another phone is released with the processor clocked high just so it can claim to perform faster?
68. Jaeel (Posts: 24; Member since: 18 Apr 2013)
The another phone wouldn't be clocked higher just for the benchmark, but also for your everyday apps.
34. ThinkDifferent (Posts: 6; Member since: 31 May 2013)
Galaxy S 4:
What is the next ???:))
40. trickster_qc (Posts: 80; Member since: 18 Jun 2013)
Every manufacturer cheats in their own way.
Samsung with the benchmark test.
Apple with the antenna gate
Motorola with the real autonomy of their batteries
Every single business in the world is telling lies in order to sell more merchandise and make more profit.
Samsung isn't different from any other.
43. yudi.nemesis (Posts: 1127; Member since: 14 Mar 2013)
Oh so thats why Exynos can have great score on benchmark but there is no advantage in real life performance.
i guess thats why Snapdragon become the king of mobile processor less gimmick and BS, just solid proof in performance with great battery and thermal efficiency.
51. Bootutu (Posts: 184; Member since: 11 Jul 2013)
Nothing to see here, move along, It had the best hardware at that time and blew everything out of the water, it was clearly stated that it runs 533MHz, so stop hating because the platform/phone you support is inferior and scores less in benchmarks. 480MHz is to save you some battery juice and still run your games smoothly, simple logic
55. juandante (Posts: 285; Member since: 23 Apr 2013)
If they downclock it to save battery anywhere else I don't see where is the cheating. If theire GPU is basically at 533Mhz, how can we call this 'overclocking' because the basic speed is 533.. No cheating, just good logic.
76. Napalm_3nema (Posts: 848; Member since: 14 Jun 2013)
It's only 533MHz in some benchmarks. Nowhere else does it perform at 533MHz unless you run a custom kernel. It's RATED for 533, but performs at 480 for everything but benchmarks.
81. Taters (Posts: 2661; Member since: 28 Jan 2013)
Nothing stopping them from releasing lines of code that allows other apps to run it at 533. Of course you rate it the highest for marketing even if you decide it is best to not go that high for thermal and battery life reasons.
Honestly, there is nothing wrong with it since everyone does it. It is just that these morons like believing Samsung is the only company doing this.
54. whoyton2 (banned) (Posts: 142; Member since: 19 Jun 2013)
Oh dear all these pointless fanboys having there day in the sun .....idiots!
56. bigstrudel (Posts: 518; Member since: 20 Aug 2012)
Well they had to come up with something to wow people because it certainly wasn't their latest iteration of Touchwiz.
59. darkskoliro (Posts: 953; Member since: 07 May 2012)
Pay students to talk behind HTC's back, trying to "fake" scores...this is why I will never buy a samsung
66. AfterShock (Posts: 2564; Member since: 02 Nov 2012)
Just more dirt and shame, not surprised.
Looks like a cheat, feels like a cheat and can't be used for your own use, it's a cheat.. yeah this reeks of deception and shame.
71. someones4 (Posts: 618; Member since: 16 Sep 2012)
I don't get it. Is overclocking a gpu considered a bad thing?
79. darkskoliro (Posts: 953; Member since: 07 May 2012)
check the Anandtech overview on this case. Once you read it you will get why this is such a cheap move
72. roscuthiii (Posts: 1785; Member since: 18 Jul 2010)
My initial reaction is that this is a pretty underhanded move by Samsung. Then I thought about all the under underhanded moves by other companies, and I'm left with this thought:
"Dear God, we're all gullible, aren't we? They all just keep nickel n' diming us, bending us over, lying to us... And we've got people here defending their favorite offender's honor? Man, they all certainly got us by the short hairs, don't they?"
73. xtremesv (Posts: 193; Member since: 21 Oct 2011)
What's the big problem? The same story happens with desktop GPUs, one should not trust synthetic benchmarks as they can be easily cheated. This alterations are necessary for benchmark addicts in order to satisfy them. There are people who don't even play games but cherish benchmarks. If you want to measure graphics performance, play and experience it for yourself.
77. deathgod (Posts: 119; Member since: 23 Nov 2011)
So someone clarify this for me: Benchmarks are used to show what the phone is capable of at its most powerful right. So how are the results cheating (serious question)? It is shady that Samsung would pull something like that when most of it's phone users won't ever be able to get that sort of scores from their phones. How big of a difference are we looking at between benchmarks and actual GPU performance?
80. Taters (Posts: 2661; Member since: 28 Jan 2013)
Are people really this fricking retarded or is this butthurt HTC, Sony, and Nokia fans?
This s**t is common sense and everybody does it but even worse. Apple is actually the worse. They market a dual core 1.2 GHz, 1 GB of ram, a below 720p screen, and a 1400 may battery as high end. How do they achieve this? Through IOS being so closed and IOS is just lines of code like this bench boost.
If you want to catch someone red handed. Look at the Xperia Tablet Z, Xperia Z, and HTC Droid DNA. This is common sense. EXACTLY the same specs as the Nexus 4, S4 Pro, 2gb ddr2 ram, similar battery sizes. So how does the Xperia tablet Z, xperia z, and droid DNA absolutely crush the Nexus 4 in benchmarks while having the exact same internals while pushing a lot more pixels? Only answer is cooked benchmarks. They all have the exact same brains and only difference is screen resolution yet they all vary in benchmarks. That should tell you something. Geez people here are so stupid.
The GS4 has the exact same cpu/GPU as the HTC One and Lg Optimus G pro but it is clocked higher and has faster ddr3 ram. Plus the bigger battery will allow the GS4 more power wasting leeway. That is all you need to know. There is no magic tricks here. Hardware is hardware. Qualcomm doesn't make a special cheating Qualcomm s600 for Samsung and a non cheating, upstanding Qualcom S600 for HTC, nor do they make non cheating S4 pros for Sony and cheating ones for Samsung.
Specs are specs. You can't lie about them.
82. Taters (Posts: 2661; Member since: 28 Jan 2013)
The GPU is rated at 533 because it can...um...reach 533. If they were being lying and underhanded, they wouldn't be able to reach 533, not even in benchmarks.......
You mean Samsung is not letting us use the full power of the Exynos 5 Octa? How dare they! They should just clock everything as high as possible and give us the choice to have a phone that lasts 10 minutes before recharging but runs the GPU at 533 and run all 8 cores at once. Those bastards....Shamesung
87. Ronny666 (Posts: 108; Member since: 08 Jun 2013)
Not suprised at all but how come only now people knew about this? Oh wait, are you guys thinking that only Samsung does benchmark optimisation and other don't? AndreiLux (Andrei F) is the person who discovered this benchmark optimisation in GT-I9500 and also a top developer in XDA forum. By the way, he owns the GT-I9500.