Global Top10 Best Performing Android Phones, December 2023

Posted: , by:Antutu     page views:   Click to copy the title and link to this article

* 1 AnTuTu was developed based on the unique technologies of different platforms, cross-platform products may not be directly comparable.

* 2 The above list, only the popular models will be displayed, monthly update.

* 3 The score displayed is the average of all data for that device, not the highest score for that device.

* 4 The list is global data but does not include the Chinese market, for more information about Chinese market, please visit our Chinese website.

No.1:iQOO 12

Average score: 2,086,638


No.2: Mi 13

Average score: 1,559,604


No.3: Galaxy S23 Ultra

Average score:1,551,689


No.1:realme GT Neo5 SE

Average score: 1,184,497


No.2: Poco F5

Average score: 1,131,237


No.3: Redmi Note 12 Turbo

Average score: 1,128,470


Only the top 10 models with AnTuTu scores are listed here, and more models' Ranking info can be found on the "Ranking" page.

The same device in the Antutu V10 will generally score higher than the V9. One reason is that the new 3D test scene is more stressful, corresponding to higher scores. Another reason is that there are more video compatibility and codec test items in the UX test, which will also make the score higher.

Here is just to explain the reason why the score is generally higher. Please note that AnTuTu V10 scores cannot be compared to V9, and the device scores can only be compared between the same version of AnTuTu Benchmark.

In addition, in order to deliver stunning image quality and precision comparable to that of the PC, AnTuTu has created two new 3D test scenes based on Unreal Engine 4. To make the output colors more realistic and detailed, we uniformly use FP32 precision in the new scenes. However, due to the higher number of FP16 units in GPU chips released in earlier years, and limited by earlier chip manufacturing processes and technologies, GPUs will combine two FP16 units to be used as an FP32 unit when performing FP32-precision calculations, and such calculations need to consume more additional computational resources, which results in poor performance on new test scenes with these GPUs, and this is more obvious on GPU chips before Mali-G710 (not included).