Google Tensor chipset; Finally, Google last night confirmed the use of the company’s proprietary 6-pixel series chipset. Although the company already has experience with chips such as the Pixel Visual Core and the Titan M security chip, this is the first time Google has entered the field of chips called SoC for smartphones.
As expected, Google has paid special attention to image processing and machine learning capabilities to build the tensor chipset. In any case, there are many questions about this chipset and its importance, and we will try to present two views on its high importance and non-importance.
Why is Google Tensor chipset important?
Primarily, Tensor is a custom chipset designed by Google to deliver the best performance in the company’s defined tasks. This means that we should expect an increase in speed and features related to image processing, audio processing and other capabilities based on machine learning. Although Google always offers a lot of possibilities in this regard, using the cloud processing and capabilities of current chipsets, but this chip can double their attractiveness.
Overall, processing speed is no longer the most important feature of a chipset. Heterogeneous calculations and performance improvements, on the other hand, may not be more important than speed, but they do play an important role in software features and phone differentiation. As Google exits the Qualcomm ecosystem and selects the necessary components, it is gaining more control over the realization of its smartphone vision.
Qualcomm has to meet the needs of a large number of companies, but Google has a much more specific vision in mind. Just like Apple, which has been producing proprietary chipsets for years to provide a unique user experience, Google intends to make its phones even more distinctive.
Also, with a dedicated chipset, the company could possibly extend the software support time of its handsets longer than before. For many of its handsets, Samsung offers 3 years of major Android updates and 4 years of security updates, and Google has taken the same approach for the Pixel 5 and previous handsets. We have to see if the Pixel 6 phones can receive software updates for a longer period of time thanks to the proprietary chipset and later generations.
Why is this chipset not so important?
If you think this chipset will be able to outperform its competitors by a long way, you probably need to prepare yourself for disappointment. Google has not released any information or benchmarks of the processor, GPU and other parts of this chipset. However, without the use of CPU architectures, the company has undoubtedly used Arm-ready designs such as Cortex-A78 processor cores. Details about the 5G capabilities of this chipset have not been announced yet.
Although Google has not yet announced by which company this chipset was produced, but according to reports, Samsung seems to be the manufacturer of Google’s own chipset. The head of Google’s hardware department also announced that this chipset will have a very competitive performance in terms of CPU and GPU performance, although the manufacturers of all chips usually make such claims.
On the other hand, it should be noted that Google’s actions for features based on machine learning and image processing are not extraordinary. In this regard, various companies have entered into action and regularly implement very important features. Google has announced that the chip could provide more advanced image processing for multiple cameras and videos.
Such features are not new features even for 2020 phones, let alone phones that will hit the market in late 2021. Examples include the Snapdragon 865 and 888 chipsets, which can use multi-camera data simultaneously, as well as effects such as HDR and bokeh when shooting 4K videos at 60 frames per second. Of course, Google may have been able to significantly improve the quality of these features, so we will have to wait until the release of the Pixel 6 phones equipped with the tensor chipset.
Are Google phones unique with a dedicated chipset?
The CEO of Google said that the beginning of the development process of this chip dates back to 4 years ago. Google started this project when features such as artificial intelligence and machine learning on mobile were relatively new. The company has always been at the cutting edge of machine learning technology and has produced chips such as Pixel Visual Core and Neural Core for its handsets due to the hardware limitations in this area.
The Google Tensor chipset is an important step in differentiating the company’s handsets, and it remains to be seen whether the Pixel 6 handsets will be able to deliver exciting capabilities thanks to the proprietary chipset. On the other hand, Qualcomm and other companies have not joined hands in the last 4 years.
Machine learning, advanced image processing, and heterogeneous computing capabilities are now at the heart of all major chipsets on the market, and these features are no longer limited to flagship chipsets. Finally, we need to see if Google’s attempt to build the chipset was done solely to show off the company’s capabilities, or if it was able to bring more attractive and useful features than other chips.
