Shulou(Shulou.com)11/24 Report--

Thanks to CTOnews.com netizens and soft media users for 1392612 clue delivery! CTOnews.com, October 22, Youtube blogger @ Mrwhosetheboss (Arun Maini) recently reviewed Google Pixel 8 Pro. In his latest video, he pointed out some of the advantages and disadvantages of the Pixel 8 series.

CTOnews.com notes that Maini's biggest criticism of the Pixel 8 series is its Tensor G3 processor. Maini points out that new generative AI functions, such as AI Wallpaper and Magic Editor, require continuous networking because they require more processing power than Tensor G3 chips can provide.

Maini explains: "for all these generative AI things, for anything that actually needs to be created with AI, such as AI Wallpaper production, Magic Editor requires a continuous network connection. (this is) because everything you do needs to go through Google's server. They are so slow that you are constantly reminded that they are not running on the device. It really makes you realize that the Tensor G3 chip in this phone is not really a flagship. "

Maini's observation runs counter to Google's marketing promotion of Tensor G3 chip capabilities. Google called the Tensor G3 a "AI first" chip in its official blog and claimed that it would bring Google's latest AI research directly to the latest phones: Pixel 8 and Pixel 8 Pro. Google also said that every major subsystem in the Tensor G3 chip had been upgraded, paving the way for generative AI on the device. It includes the latest generation of Arm CPU, upgraded GPU, new ISP and Imaging DSP, and the next generation of TPU specifically designed to run Google's AI model.

However, Google's claims about the importance of the Tensor G3 chip and what it can accomplish seem to conflict with the way Pixel 8 Pro actually handles generative AI tasks. Google hinted in its blog that TPU (Tensor processing Unit) plays a leading role in handling AI tasks, while ignoring the key role played by Arm's CPU and GPU (as used in Tensor G3) in performing device-side AI tasks.

According to Arm itself, its CPU and GPU play a central role in handling artificial intelligence tasks: "as AI computing moves from the cloud to data collection sites, Arm CPU and MCU (microcontroller unit) technologies already handle most AI and ML workloads at the edges and endpoints. CPU is the core of all AI systems, whether it fully handles AI or works with coprocessors such as GPU or NPU to handle certain tasks."

It is worth mentioning that Google also tried to prevent reviewers from installing popular benchmark apps in the Play store before lifting the ban. Testing of Geekbench 6 shows that although the Tensor G3 has a fairly new CPU architecture, its performance is closer to the mid-range Qualcomm Snapdragon 7 + Gen 2 than its current flagship chip Snapdragon 8 Gen 2.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.