The Google Tensor from Pixel 6 lags behind in the performance test

Many fans of Android devices have been looking forward to the introduction of the new smartphones from the Google Pixel series in recent times. This is the news about Google Pixel 6 and Pixel 6 Pro. Perhaps the most interesting feature of the recently introduced innovations is the implementation of in-house processors. It’s about Tensor chipsets. GSMAren’s website has now revealed something about the rough performance of Google’s new home processor.

Google Tensor consists of an interesting combination of individual cores

Although the deployment of custom chipsets in the new Pixels has indicated much better optimization and very high performance to many, Tensor processors don’t seem to be doing as well as chips on iPhones. Unusually, the tensor consists of a pair of Cortex-X1 cores, two Cortex-A76 cores from 2018 and a four Cortex-A55.

The first mentioned cores have become very popular this year. The Cortex-X1 is the primary performance core in both the Snapdragon 888 and Exynos 2100 chipsets. So we have some idea of ​​the performance it could offer. Google Pixel 6 and Pixel 6 Pro are, of course, in single-core performance very comparable to competing chips.

In a multi-core test, the competition leads. However, Google has an opportunity to improve

However, when it comes to the multi-core test, the difference is very noticeable. Due to the special combination of individual cores, the overall performance using multiple cores is relatively low, which is an unpleasant surprise. Both the Snapdragon 888 and the Exynos 2100 are better processors in the rough performance test. The performance of the chips used in the devices of the American giant Apple is significantly higher.

On the other hand, the graphics performance is reportedly on par with other flagships. In the case of the Tensor processor, the Mali-G78 MP20 became the graphics accelerator. However, as you probably know, gross performance is one thing. The second is optimization, which Google undoubtedly has great. The chip also has additional hardware for machine learning. It will also please the image processor, which makes the news good photo cars. We will see how Pixels will do in practice. We believe that Tensor is a bit better than it may seem at first glance from benchmarks. Software updates can sometimes work wonders.