Here's what Google has to say about Tensor G2 benchmark performance

1 year ago 1012

Here's what Google has to say about Tensor G2 benchmark performance

The heart of Pixel 7 and the Pixel 7 Pro smartphones is Google's own Tensor G2 chipset, a 4nm chipset with 2 Cortex-X1 cores, two Cortex-A76 cores and four Cortex-A55 cores. The new chip runs on essentially the same core layout as the original Tensor, but has a more efficient 4nm process. Built upon a 4nm manufacturing process, it utilizes three core (2+2+4) clusters, Tensor Processing Unit (TPU) for machine learning and thr same Titan M2 security co-processor. Software-based Face Unlock is there, along with an in-display fingerprint scanner.



Pixel 7


Pixel 7 Pro

globally earlier this month. Recently, Alphabet and Google CEO Sundar Pichai thanked positive reviews of the newest series and said that the company had its highest selling week ever for the Pixel 7 smartphones. Both these phones come with Google Tensor G2 – the second generation of the company’s proprietary silicon – that allows it to do “what it could not with other chipsets.” Now, a senior executive has cleared Google’s stand on benchmark performance of the silicon.
While we are yet to review the phones, reports by various tech publications suggest that although Pixel 7 is a great smartphone to easily carry out daily tasks, the Tensor G2 chip under the hood lags behind its competitors in terms of benchmark performance. But Google is comfortable with Tensor not winning benchmarks.

Monika Gupta

, senior director of product management for

Google Silicon Teams

, says that the company wants to focus on what the company needs "five years from now” for its chips.

"I’m not making decisions based on where machine learning is today, and I can say that because I work at Google. I know where the software team wants to take the user experiences five years from now. That’s the benefit of not being a merchant silicon supplier, but an in-house silicon supplier," Gupta said during the Made by Google podcast.
She also explained why Google does not discuss the raw firepower that Tensor offers and what the company's priorities are. "I think classical benchmarks served a purpose at some moment in time, but I think the industry has evolved since then. And if you look at what Google is trying to do by pushing AI innovations into a smartphone, because we feel like this is the approach that will deliver helpful experiences like some of the ones I just mentioned, classical benchmarks were authored in a time where AI and phones didn’t even exist," Gupta explained.
As per the executive, what Google benchmarks is the actual software workload that is running on the chip. She highlighted that both Pixel 6 and Pixel 7 series of smartphones come with "amazing innovations" that use AI and machine learning (ML) in carrying out tasks. For example, Magic Eraser that removes people from the background and Photo Unblur that uses ML algorithms to remove blur from any photo.


FacebookTwitterInstagramKOO APPYOUTUBE

Read Entire Article