Phones Samsung tweaking benchmark scores with an overclock when running benchmark apps?

[lazard]

Ahh... The "other people do it" excuse. They cheated. Obviously it's a big deal to Samsung, or they wouldn't have put in the extra work to do it.

The Android users who care about benchmarks pretty much all already run custom kernels which overclock their GPU anyways (which is what Samsung did).

 

[AppleFan1984]

Really?

The footnotes at the bottom of that apple.com page include:

Based on theoretical peak speeds. Actual speeds will be lower.
Battery life varies by use and configuration

There are many more.

Most advertised specs reflect optimized conditions. The details discussed at anantech.com are no doubt of great interest to that audience, but I'm not so sure this warrants much controversy.

 

[lazard]

The Exynos Octa uses a PowerVR SGX 544MP3 GPU which has a max clock of 533 MHz. In Anandtech's tests they stated that the GPU clocked in at 532 MHz. So technically it's not even an overclock.

 

[Brandon263]

I wonder whether this will affect future Samsung reviews.

Original Source: Definitely worth a full read!


----

Synthetic benchmarks are never a good indication of real world performance on any handset IMO but it appears Samsung is being shady when it comes to Android benchmark applications.

http://www.slashgear.com/samsung-ga...tuned-for-benchmarks-research-spots-30292185/



A very sneaky move from them IMO.

Benchmarks are okay for some but the way I've seen them gamed over the years by ATI and Nvidia over the years on PC (Quack3 Anyone?) makes me think of them as a more secondary metric over other aspects (design, features, price).

Peoples' thoughts on this? Are Samsung being underhanded or is there a legit reason for higher clockspeeds when running a benchmark app?

 

[Technarchy]

Interesting...

Ramping up full system power to cheat the benchmarks, and then ramping down during actual performance.

This might explain why many have noted the Exynos GS4 is laggy compared to the Snapdragon 600 even though the benchmarks suggest otherwise.

Also suggests Samsung is losing their SOC mojo to Qualcomm, which is something I never would have expected because the Nexus S, and GS2 were monsters that smashed everyone else's contemporary products.

Now I have to question the GS3 and GS4's performance claims from a technology standpoint.

 

[unobtainium]

Android flagship phones are, and will be, noticeably faster than any iphone

Nope. They are, and will be, faster on paper. In real world use they'll be about the same, often laggier. Not "noticeably faster" and, if anything, the opposite. Numbers are absolutely meaningless. It's embarrassing that Samsung cares so much about numbers that they're willing to actually tweak the S4 to perform differently when running benchmarks.

 

[Oletros]

Some people take synthetic benchmarks seriously (hence the necessity for Samsung to game the system like they have).

If they weren't at all important, I doubt that "benchmark booster" code would even exist.

I've seen the same thing over the years from Nvidia, ATi, 3dfx and others so this is nothing new (but most of that was driver software tweaks rather than hardware overclocks).

It is funny to see the smartphone works of benchmarking emulating what PC component manufactures were doing years ago.

Well, even companies are optimizing the rendering and javascript engines to do better in benchmarks like Sunspider

 

[Reason077]

Samsung caught cheating on Galaxy S4 benchmarks

Apparently, the Galaxy S4 contains code to "optimize" (by over-clocking itself) when a benchmark app is detected, artificially inflating its results.

http://www.anandtech.com/show/7187/looking-at-cpugpu-benchmark-optimizations-galaxy-s-4

Fair play by Samsung, or deceptive behavior?

 

[lazard]

Apparently, the Galaxy S4 contains code to "optimize" (by over-clocking itself) when a benchmark app is detected, artificially inflating its results.

http://www.anandtech.com/show/7187/looking-at-cpugpu-benchmark-optimizations-galaxy-s-4

Fair play by Samsung, or deceptive behavior?

Exynos Octa GPU = PowerVR SGX 544MP3

PowerVR SGX 544MP3 spec = up to 533 MHz

clock frequency of PowerVR SGX 544MP3 during Anandtech's runs = 532 MHz

532 MHz < 533 MHz = not overclocked

 

[Oletros]

An update from Anandtech:

http://www.anandtech.com/show/7192/update-on-gpu-optimizations-galaxy-s-4

It is interesting that quote:

Since our post yesterday we've started finding others who exhibit the same CPU frequency behavior that we reported on the SGS4. This isn't really the big part of the discovery since the CPU frequencies offered are available to all apps (not just specific benchmarks). We'll be posting our findings there in the near future.

 

[ross1998]

Doesn't really matter-

1. Benchmarks don't prove nothing and are pretty useless.

2. Samsung targets top of the line android phones with benchmarks not iPhones.

 

[cynics]

Doesn't really matter-

1. Benchmarks don't prove nothing and are pretty useless.

2. Samsung targets top of the line android phones with benchmarks not iPhones.

I agree. Benchmarks are for those who want to increase the speed of their device and be able to compare it. A before and after.

People focus too much on comparing them to other devices or even across platforms with a totally different OS.

 

[Brandon263]

Samsung has responded (rather implausibly) to Anandtech:

http://global.samsungtomorrow.com/?p=26314

Under ordinary conditions, the GALAXY S4 has been designed to allow a 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. 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.

The maximum GPU frequencies for the GALAXY S4 have been varied to provide optimal user experience for our customers, and were not intended to improve certain benchmark results.

Samsung Electronics remains committed to providing our customers with the best possible user experience.

 

[Michael Goff]

Even if they did, I don't care. My Nexus scored horribly on benchmarks, still go it. People who buy because of benchmarks kind of get what they deserve on this one.