What a cover up! Google prohibits Pixel8 from installing testing software: poor performance of self-developed chips

On October 15th, KuaiTech reported that recently Google released its latest flagship smartphones, Pixel8 and Pixel8Pro, both equipped with its latest self-developed TensorG3 processor.

TensorG3 is based on a 9-core CPU architecture, including one Cortex-X3 super core with a main frequency of 3.00GHz; Four Cortex-A715 cores with a main frequency of 2.45GHz; Four Cortex-A510 small cores with a main frequency of 2.15GHz.

The GPU uses a 10 core ArmMali-G715GPU, which supports hardware level ray tracing acceleration capabilities like the Apple A17Pro.

However, recent news has shown that Google has taken extraordinary measures against Pixel8/8Pro, and all new users of Pixel8 after its release are unable to directly download benchmark testing tools, including GeekBench and 3DMark, through the app store, in an attempt to suppress the spread of scoring results.

GeekBench is a commonly used software for testing and scoring CPUs, which can intuitively obtain single core and multi core scores. 3DMark is a testing software for evaluating GPU performance. In the vast majority of mobile phone evaluations, these two testing software are important reference criteria.

Google is very aware of the relative performance shortcomings of TensorG3 chips and emphasizes that the chip is designed with efficiency and artificial intelligence as its goals, rather than simply pursuing performance or scoring. It is necessary to mention a sentence from Lei Jun here: "Although running points cannot fully represent the mobile experience, running points are definitely the foundation of the mobile experience

Although Google prohibits users from installing these testing software, there are still ways to bypass this restriction, so TensorG3's Geekbench and 3DMark scores have still been exposed.

One of the Geekbench6 results showed that the TensorG3 chip on the Pixel8Pro phone scored 1760 in a single core test and 4442 in a multi-core test, which is far from the Qualcomm Snapdragon 8Gen2 chip.

Efficiency testing also shows that TensorG3's GPU efficiency ranking is also relatively low, comparable to older chips such as Snapdragon 888.

Google's approach is quite reminiscent of Jin people hiding their ears and stealing information.