Home Ā» Google: “Very satisfied” with the low benchmark scores of the Tensor G2 chip, AI functions are more important than traditional performance scores

Google: “Very satisfied” with the low benchmark scores of the Tensor G2 chip, AI functions are more important than traditional performance scores

by admin
Google: “Very satisfied” with the low benchmark scores of the Tensor G2 chip, AI functions are more important than traditional performance scores

Google has included its latest Tensor G2 chip in the latest Pixel 7/Pro series of phones. With the help of the Tensor G2 processor, the Pixel 7 and Pixel 7 Pro will bring more useful personalization features for photos, video, security and voice recognition, Google said.

However, the chip’s actual performance test performance is not excellent. Geekbench 5 single-core running score is 1068, multi-core running score is 3149, which is only a small improvement over the previous generation Tensor chip, and it can’t keep up with the level of the flagship chip in 2022.

Google responded to the performance of the Tensor G2 chip, saying that they value the AI ā€‹ā€‹ā€‹ā€‹function of the chip more than the hardware performance.

Monika Gupta, senior director of product management on Google’s silicon team, said they designed the silicon to focus on machine learning needs five years from now.

ā€œI think traditional benchmarks made sense at some point, but times are moving and if you can see what Google is doing to try to advance smartphones with AI innovation, you can understand traditional benchmarks It has fallen behind the era of artificial intelligence and even the era of mobile phones.ā€

ā€œSo for us, our benchmarking is to run actual software workloads on the die, and then try to use each generation of Tensor die to make the software better, whether itā€™s better quality, better performance, more low power consumption.”

Google has said it is “very satisfied” that the Tensor chip didn’t win the benchmark because of its focus on prioritizing the end-user experience.

See also  Chat on WhatsApp without smartphone arrives

The Google Tensor G2 chip uses Samsung’s 5nm process. From the running test, the Arm Cortex-X1 super core frequency in Tensor G2 is 2.85GHz, which is only 50MHz higher than the original Tensor’s 2.80GHz.

You may also like

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.

This website uses cookies to improve your experience. We'll assume you're ok with this, but you can opt-out if you wish. Accept Read More

Privacy & Cookies Policy