If you’re worried about the Tensor chip, you bought a Pixel for the wrong reason

Some Pixel owners love using AI and did not buy the device to play games

But here’s the thing, some of us Pixel users own our phones because we love the AI capabilities of the device and do not care about gaming on our phone. In fact, if you treasure gaming over AI, I’d question whether the Pixel is the phone for you in the first place. 

Yes, the Tensor AP cannot compete with Apple’s A-series AP, Qualcomm’s Snapdragon 8 silicon, and more

I’d be the first to admit that the Tensor APs can’t compete with top APs like Apple’s A-series APs, Qualcomm’s Snapdragon 8 Elite APs and MediaTek’s Dimensity 9500 APs when it comes to Geekbench scores, performance, and gaming. But when it comes to specific AI tasks for Pixel phones, Google can custom-build such capabilities into the Tensor chip, such as the Tensor Processing Unit (TPU).

The Tensor G5 AP powered the Pixel 10 series | Image by Google

Because the TPU is tailored for Google’s AI-related features like computational photography, real-time translation, and voice recognition, these features can run on-device instead of in the cloud. Running AI on-device allows features to respond faster, reducing latency.

G.oogle’s TPU focus on AI and Machine Learning

With the TPU focusing on AI and Machine Learning, complex tasks handled in real-time on a Pixel, such as live translation and the Magic Eraser, consume less battery power. While the Neural Engine in Apple’s A-series chip is powerful, the customized Tensor AP can help Pixels run exclusive features such as real-time voice call screening, and advanced photo processing.

Yes, Google is giving up raw CPU and GPU performance with the Tensor AP in return for some advantages in AI. And that brings us to the reason you bought a Pixel. If you’re like me, you probably bought your Pixel for AI features like the Magic Cu, which allows your phone to be proactive rather than reactive.

Having the Tensor G6 and Tensor G7 power the Pixel 11 and Pixel 12 are not disasters for Google

We also need to point out that the Tensor G6 this year, and the G7 next year, will both be built on TSMC’s 2nm process node. This matches the technology that will be used by TSMC to build Apple’s A20 and A20 Pro APs this year. The bottom line is that this year’s Tensor G6 AP and next year’s Tensor G7 AP are not the fastest and most powerful APs, but I don’t believe this is anywhere close to a disaster for the Pixel 11 or Pixel 12 lines.

Aman said that he switched from the iPhone to the Pixel and felt a drop in performance. But as he admits, he is a hardcore mobile gamer. That’s exactly the type of user that the Pixel is not made for, and had he asked me before buying his Pixel, I would have dissuaded him from making this purchase. He also seems surprised and shocked that Google would continue to use the Tensor AP on Pixel models going forward.

It really shouldn’t be shocking. The Tensor chip gives Pixel devices the AI features that I believe that most Pixel users want over better game playing. Google’s tagline could be “AI first, performance last,” but this is not the insult Aman thinks it is. If AI is more important to you, the Pixel should be your phone.