Samsung puts the focus back on the energy efficiency with the first autonomy benchmarks of the Galaxy S26which put the Pixel 10 Pro XL equipped with the SoC Tensor G5 from Google. The tests reveal clear differences in battery life even when the Samsung terminal integrates a battery approximately 1,000 mAh smallersomething that in theory should put it at a disadvantage compared to the Google device.
The results suggest that the problem does not lie in the energy capacity of the Pixel, but in the optimization level of Tensor G5 itselfwhose overall performance and efficiency have been the subject of debate since its arrival on the market. This scenario once again puts on the table an increasingly evident reality in the mobile sector: silicon efficiency and system power management weigh as much or more than battery capacity.
Tensor G5 architecture: focus on AI with limitations on CPU and GPU
Table of Contents
He Tensor G5 maintains Google’s traditional approach to artificial intelligence processing and the tasks of machine learningbut its design presents some debatable technical decisions compared to the most recent SoCs in the mobile sector. The chip uses a configuration of Eight-core CPU based on ARM architecturewith a hybrid design that combines high-performance cores, intermediate cores and energy-efficient cores.
Specific processor configuration includes 1 Cortex-X4 core high performance running at 3,78 GHzaccompanied by 5 Cortex-A725 cores intended for intermediate loads 3,05 GHzin addition to 2 Cortex-A520 cores efficiency-oriented that operate at 2,25 GHz. This distribution seeks to balance power and energy consumption, although in practice the chip does not always manage to compete with the most recent solutions from Qualcomm or MediaTek.
The SoC also integrates a Fifth generation TPUspecifically designed to accelerate AI and machine learning processesalong with a GPU Imagination IMG DXT-48-1536 architecture based PowerVR that operates at 1,10 GHz. On paper, this GPU should be close to the performance of solutions like Adreno 732/740 o Mali-G715 MP7although it lacks support for ray tracinga feature that is beginning to appear in other high-end mobile chips.
The set is completed with a Samsung Exynos 5G modemin charge of mobile connectivity. This component is part of the chip’s communications ecosystem and also contributes to the overall energy performance of the device.
Two key weaknesses of the Tensor G5 compared to the competition
The criticisms towards Tensor G5 They mainly focus on two technical factors that affect both performance and energy efficiency. First of all, Google would have chosen relatively old ARM coresalmost three years old in some cases, a decision that could be related to the optimization of development costs but that ends up being reflected in the performance against the most recent SoCs on the market.
Secondly, the GPU IMG PowerVR DXT-48-1536despite having been developed in collaboration with Imagination Technologiescontinues to show irregular behavior in demanding scenarios. Even after several software updates intended to improve your performancethe graphic section continues to lag behind other solutions present in the current high-end.
This imbalance between theoretical power and real performance also ends up influencing the energy consumptionsomething especially relevant when the device faces prolonged loads such as games, video playback or intensive applications.
The Galaxy S26 shows that chip efficiency is still key
The new ones autonomy benchmarks of the Galaxy S26 clearly reflect the real impact of these limitations. In video playback tests, the Samsung terminal achieves more than 3 additional hours of autonomy in front of Pixel 10 Pro XLa considerable difference considering that the Samsung device integrates a lower capacity battery.
The gap widens even further in gaming scenarios, where the Samsung terminal achieves almost double the duration regarding the Google model. This type of testing is especially insightful, as heavy GPU and processor usage often tests the power efficiency of the entire system.
In terms of average active use, the Galaxy S26 reaches 15 hours and 20 minuteswhile the Pixel 10 Pro XL stays at 12 hours and 29 minutes. These data reinforce the idea that Increasing battery capacity does not compensate for an inefficient processorespecially when the power consumption of the SoC becomes the limiting factor of the device.
The efficiency of the SoC is once again decisive in current smartphones
The tests once again highlight a clear trend within the mobile sector: the processor optimizationthe thermal management and the energy balance of the system They are factors as determining as the capacity of the battery itself.
In this context, the Galaxy S26 proves that a more efficient silicon design can deliver better results even with a smaller batterywhile the Tensor G5 is once again at the center of the debate about Google’s strategy in the development of its own chips for smartphones.
Vía: Wccftech