Google first launched Tensor in the Pixel 6 series late last year – a high-end semi-custom chipset with Google’s own hardware sauce on top of it.
While the Pixel 6 series hasn’t quite flaunted the best-in-class flagship performance, it still offers plenty of power and impressive machine learning capabilities. Now, the mid-range Pixel 6a also uses a Tensor chipset, but would it be better if Google made a Tensor Lite processor for the Pixel A series instead?
Arguments for a mid-range Tensor SoC
Jimmy Westenberg / Android Authority
Now, it seems a bit counter-intuitive because a high-end powerful processor is such a great thing in a smartphone. Google Tensor already fits the law and is a very capable SoC. But there are a few reasons Google might consider downgrading things a bit for future Pixel-A phones.
First, just take a look at the Pixel 6a to realize that Google seems to be prioritizing the use of the Tensor chipset above everything else. Sure, you get a high-end processor, but that comes at the cost of many other features. High refresh rate? He went. Fast wired charging? no. More competitive camera hardware? Forget about it. A newer version of Gorilla Glass? No, you just have to use Gorilla Glass 3.
Read more: Has Google lost its way with the Pixel A series?
Core processors are expensive. By choosing a less capable but potentially cheaper Tensor processor, Google could spend money and resources on these aforementioned aspects of the mid-range Pixel design. Alternatively, Google could take the cheaper route, as it did with the Pixel 4a, and lower the asking price. The Pixel 7a at $399 will be more acceptable than the $449 Pixel 6a.
Tensor’s cheaper mid-range could allow Google to spend money on other features or pass on savings to consumers.
There’s also a small argument to be made that the Pixel 6a doesn’t quite exploit the Tensor chipset in the same way the Pixel 6 series did. Sure, you have face blurring and the Magic Eraser camera as well as offline voice typing, but you’re also saying goodbye to some camera features. Like Motion Mode. There is also a lower resolution and slower refresh rate display than the Pixel 6 Pro, which requires less GPU power to drive. If the Series A is not going to take advantage of all the key features enabled by the Tensor chip, why use the full chip in the first place?
Returning to the drawing board for the mid-range processor will also allow Google to address some of Tensor’s shortcomings faster than its annual launch cycle. For example, we previously covered important reception issues for the Pixel 6 range that may now plague the 6a as well. Our Pixel 6a review found that the device was also hot. Mid-range chips, by their more economical nature, tend to run cooler and drain less battery than flagship processors. Google will likely offer a smaller battery with the same screen on time, or provide longer endurance with the same battery capacity. This also opens the door to a more compact mid-range Pixel phone, in line with the pocket-friendly pixels of the past.
More on Tensor: Comparison between Google Tensor and Qualcomm Snapdragon 8 Gen 1
In short, the released Tensor Lite chipset could bring cheaper Pixel A series phones price (or more features), a cooler chip, more reliable connectivity, and improved battery efficiency compared to the flagship Tensor SoC.
What would a mid-range Tensor look like?
Ryan Haines / Android Authority
The Tensor chipset is actually related to the Samsung Exynos family. It is designed and manufactured by Samsung and uses Arm CPU cores and Arm GPU. It even features the same modem as the Galaxy S21 series. It makes sense that any mid-range Tensor processor would share a similar foundation.
The original Google chipset opted for an eclectic octa-core CPU design, which includes two Cortex-X1 cores, an older Cortex-A76 core, and four Cortex-A55 cores. This is a high-performance setup, so Google can ditch the Cortex-X cores of the theoretical Tensor Lite processor in favor of medium and small cores (whether it’s in 4+4 or 2+6 layouts). There are some upsides to dropping Cortex-X cores. They are slightly larger than medium and small cores and require more cache for peak performance, so there is some silicon savings and licensing costs can be reduced by removing them.
It makes sense for a theoretical Tensor Lite SoC to ditch the power-hungry Cortex-X cores, but the AI silicon should still be kept.
On top of that, Arm’s Cortex-X cores are designed with performance in mind rather than battery life, which generates a lot of heat. Processors without Cortex-X, such as the Dimensity 8100-Max and Snapdragon 870, show less severe throttling than what we see in modern flagships and seem to offer solid battery life. Arm’s latest mid-core – the Cortex-A715 – will be a suitable alternative. Arm claims that the A715 can reach the same performance as the Cortex-X1, which makes it suitable to support existing flagship Pixel features. Google has shown that it is not against using legacy CPU technology. The Cortex-A77, or better yet the Cortex-A78, paired with small Cortex-A55 cores, will still provide plenty of performance at an affordable price.
More silicone coverage: What you should know about Arm’s 2023 CPUs and GPUs
Google will almost certainly use the Arm GPU in a theoretical mid-range processor, with the current Tensor SoC using the Arm Mali-G78 MP20 GPU. However, there are good reasons to reduce the number of shading kernels; Gaming isn’t a top priority in the mid-range, and GPU cores take up a lot of silicon and hence their cost. Instead, a newer mid-level Arm graphics core, such as the Mali-G610 or Mali-G615, is more efficient and power efficient. Modern mid-range GPUs from Arm are the same basic designs as the main GPUs, and differ mainly in the number of shaded cores. So while switching to mid-range graphics and reducing the number of shader cores would result in solid performance compared to the flagship silicon, it should still provide decent performance for advanced gaming.
We want Google to keep the machine learning silicon (TPU) dedicated to the proposed Tensor Lite processor, as it’s heavily integrated into the phone’s image processing line as well — it’s what gives the Pixel 6 its smart photography and smarts. By contrast, mid-range processors from Qualcomm and Mediatek typically use less machine-learning-capable hardware than the flagship silicon. But the TPU is the key part of Tensor’s identity, enabling existing Pixel features and tangible benefits like offline voice typing, complex camera features, and cheaper live voice translation on the Pixel.
Mid-level silicon isn’t just about performance and cost, but it could also free Google to build a more compact device.
The modem, however, is an area ready for improvement. The original Tensor uses an external 5G modem to support some of the best features, but an external modem usually consumes more power and has more space than a built-in modem. Moving to a less capable but integrated modem will save on power and component costs, but with some drawbacks for peak speeds and future 5G. However, do you really want 10Gbps speeds and other marginal features instead of better battery life and lower costs in a mid-range phone?
Further reading: Best cheap phones of 2022
Either way, a mid-range Tensor SoC with a less impressive CPU, scalable GPU, and modem may result in a design that may be less capable than the original Tensor. But these cuts will free up the silicon area on the Tensor Lite chip, resulting in a smaller design with better battery life and fewer concerns about overheating. Aside from the obvious benefits of a phone that runs longer and cooler, it may also allow Google to try new things in terms of phone design while lifting battery and cooling restrictions somewhat. That could mean a slimmer design, a more compact phone, or something like a foldable clamshell. It could also lower manufacturing costs, as more chips could be produced from the same silicon wafer.
The benefits outweigh the negatives
This does not mean that the mid-range Tensor processor for the Pixel A series will be without any flaws, as there are some advantages to using a high-end SoC instead.
Continuing to use the previous year’s flagship Tensor processor gives the Pixel A phones a lot more power and better performance than their competitors. This power ensures overall smooth performance and a seamless experience when playing advanced games, ensuring Google’s bragging rights similar to Apple’s iPhone SE. The flagship Tensor chip in the Pixel A phone is also simplifying things for Google when it comes to moving key Pixel features to the Pixel A series down the line, as well as streamlining the development process to keep its phones up to date.
As we found with the iPhone SE, the use of flagship silicon in a mid-range smartphone can be overstated.
There is also the question of whether Google will save any money by switching to the mid-range Tensor chip. It is possible that the company will use the flagship Tensor chipset because it already has a lot of inventory. In addition, the new chip requires additional costs in terms of research, development and start-up of production.
However, there are clearly significant benefits to using a mid-range Tensor processor. Between a potentially cheaper price, a more improved mid-range feature set, a cooler chip, and a more battery-friendly phone, the gains are clear. Plus, mid-range doesn’t have to mean it’s running low on power, as today’s best mid-range processors can take the fight to older flagship SoCs.