Did Samsung Cheat on Us? Samsung Galaxy Note 3 with AnTuTu B
Recently, the GPU of Samsung Galaxy S4 which powered by Samsung own double quad-core Pro Exynos 5 Octa 5410 has been exposed to overclock specially aiming at benchmark Apps in order to get an impressive score. However, Samsung denied about the fraud lately.
Samsung's statement then was nearly identical: "the maximum GPU [graphical processing unit] 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."
However, the company also contradicted itself and confirmed what the releasing media had reported about specifically juicing certain benchmarks within the same statement, adding:"A maximum GPU frequency of 533MHz is applicable for running apps that are usually used in full-screen mode," Samsung said, "such as the S Browser, Gallery, Camera, Video Player, and certain benchmarking apps, which also demand substantial performance."
Nevertheless, Whether Samsung had cheated on benchmarking score or not, or should we trust what they had said? In response to this phenomenon, AnTuTu has released a brand new Antutu Benchmark X Editor which is particularly modified to protect the software from being recognized as benchmark. Here, we have a benchmarking test with another Samsung hot flagship device-Galaxy Note3 .
The Note 3 we used is N9005 powered by 2.3GHz Snapdragon800 CPU and Adreno 330 GPU, with 3GB RAM and 32GB ROM, and running Android 4.3.
1, General use
With main interface and no background task, one kernel was adopted and the frequency was down to 300MHz while sometimes could arrive 1.7GHz. If there are more tasks, the kernel will unlock one by one.
2,With AnTuTu Benchmark
After running AnTuTu Benchmark, it showed that the device opened 4 kernels directing at AnTuTu kind Benchmark and all the kernels hold with the highest frequency 2.3GHz. But in additional, we installed AnTuTu Benchmark X Editor which could not be recognised as benchmark app but a ordinary App. The device still increased the performance output step by step according to the normal procedure.
3. Benchmark detail
A, Dalvik test
The images above may not so clear. The left one is full open 4 kernels with 2.3GHz running standard benchmark, and the right one is only 2 kernels with 2.3GHz and 1.7GHz.
B, CPU test
In CPU test, the gap between the two versions is somehow big on CUP integer and CPU floating point. When running CUP integer, both are 4 kernels full open and the frequence up to 2.3GHz. While with CPU floating point, the frequence of standard benchmark maintained 1.0GHz to 1.7GHz, whereas the X version was drop to 1.0GHz-1.3GHz. This kind of test lay not so much load on hardware.
C, RAM operation and speed test
The two tests with standard version kept 1.0GHz. The X version unlocked 4 kernels and maintained 1.3GHz at the beginning, after the device detected no need for so much performance, 2 kernels had been shut down.
D, Multitask test
In Multitask test, the standard version backed to a full open 2.3GHz, and ranged from 1.0GHz to 2.3GHz. On the other hand, the X version unlocked 4 kernels and maintained at 1.5GHz. In testing, the number of unlocked kernel was changeful.
E.2D Graphic and 3D Graphic test
2D graphic mainly tests GPU performance. In standard version, the device was on a situation of 4 kernels unlocked with 2.3GHz. While only 1 or 2 kernels was unlocked in X version ranging from 300MHz to 1.5GHz. The 3D graphic test is same with 2D graphic.
So, how different the final outcome between the two versions is?There was about 1000 points between the two versions- In standard version the Samsung Galaxy Note 3 got 30599, and in the X version it got 29741. 1000 points only took 3.3% in the total score. This fluctuation is within the normal fluctuation range.
We can see clear that Samsung has set a performance full open mode on AnTuTu Benchmark . However,during the test the device did not over the highest frequence 2.3GHz of Snapdragon 800 just keeping a performance full open mode. This is not cheat.
AnTuTu has always though that basing on hardware performance, our users need a device with high-end performance, while Benchmark is the tool to prove those things. Therefore, many smartphone makers take advantage of the psychology of this then cheating on Benchmark. We AnTuTu do not oppose those makers adopting the highest GHz ,because this is what benchmark test about-- to reflect the max performance of a hardware. BUT,you should let all the apps operating on your devices have this treatment not benchmark only. If you have run both standard version and X version, you would find that there are over 10% gaps in general.
AnTuTu Benchmark is born for the Users to test the devices. Nowadays, many chipsets makers like Samsung,Huawei and Broadcom all have plan to develop a own-made benchmark. And Qualcomm has released Vellamo to show its performance better than other rivals. Yet, the Benchmark developed by chipset-maker themselves may be only suit for their own chipset. Who knows?