We aren’t even to the point where Google has officially announced plans to bring their Tensor SoC to Chromebooks, but between leaked info, common sense, and gut feelings, I’d say a desktop version of Tensor is a foregone conclusion at this point. Reports put the arrival of Google’s custom Chromebook SoC sometime in 2023, so there’s time for things to shift and change, but I’d recon we’ll see a new Pixelbook arrive with Tensor inside at some point and I’m very, very excited by this.
I feel like we’re about to see what Chrome OS-specific ARM SoCs are capable of as we welcome MediaTek’s Kompanio 800 and 1000 series into the fold soon. Both of these chips were designed to be used for Chrome OS, and with MediaTek’s long-standing relationship with Google and Chromebook manufacturers, I’m expecting some pretty sweet performance. These aren’t off-the-shelf SoCs crammed into a Chromebook: they were built for Chrome OS.
And that’s what has me so excited for Tensor on a Chromebook. Google is busy building out this SoC that will be tuned specifically for the operating system that they have 100% control over. Remember the wild increase in speed and power Apple achieved with the original M1 chip? That sort of performance could be on the table with Tensor for Chromebooks, and I think there’s good reason to believe Google might share the wealth a bit.
Waiting on a Pixelbook
My first reason for thinking this is the fact that we have pretty firm confirmation that there’s no new Pixelbook on the immediate horizon. Two years after launching the Pixelbook Go, we’d fully expect Google to deliver on a new Google-made Chromebook experience, but that doesn’t seem to be happening. At least not yet.
With Tensor for Chromebooks, everything is a bit of a guess at this point. Though we have good reason to believe this project is in the works, we don’t have any true confirmation and we also don’t know for sure that Google is 100% making another Pixelbook. With Tensor on the way, it would only make sense to create a new Pixelbook to alongside it, but again, that’s just using common sense and a bit of logic to get to that conclusion.
However, if we can all agree that this is very likely happening at some point in the future, it begs a very interesting question: why wait? It’s been 2 years since the last Pixelbook came out and it could be another 2 years before the next if the assumed 2023 timeline is correct. In that time, Google surely could deliver an updated Pixelbook or Pixelbook Go without too much fuss, right? I’ve begged for this in the past and if Google’s goal with the Pixelbook brand was to really compete in the Chromebook market, they’d be foolish for not at least putting out an iterative update to one of the Pixelbooks.
But I’d wager that isn’t their goal with the Pixelbook lineup at all. Think about it: every Chromebook Pixel or Pixelbook has arrived to usher in some sort of new normal for Chromebooks. The original Chromebook Pixel was created to show manufacturers what a premium Chromebook could look like. The 2015 followup arrived to show off the new battery prowess of Intel’s latest chips and to introduce USB Type C to the Chrome OS world. The Pixelbook paved the way for well-made, thin, light convertible Chromebooks with pen support and the Pixelbook Go was the poster child for simplicity + quality in a more affordable, thin, light package. And we can’t forget the Pixel Slate fully introducing the world to what Google thought was going to be another run at tablets.
The point? Google’s own Chromebooks serve a specific purpose in the industry. Google’s not building these devices to directly compete in the space. Instead, they are making them in order to be more of a lighthouse, showing the way to manufacturers and giving them a solid, concrete example of what can be done next with a Chromebook.
With that in mind, let’s revisit the idea of Google building some sort of interim Pixelbook while we wait on Tensor. What would the point be? If Google uses their own Chromebook hardware to highlight a new form factor or new hardware addition to the Chrome OS ecosystem, what new thing would they be showing off with an iterative update to the Pixelbook or Pixelbook Go? Other than just making a slightly improved version of those devices, what purpose would those Chromebooks serve?
If being a lighthouse for manufacturers is the purpose of Google’s own Chrome OS hardware, it makes a lot more sense for them to wait and build a new Pixelbook when they can put Tensor inside. THAT is something new, something ground-breaking and worth a new Pixelbook to usher it into existence. Until then, however, I just don’t see Chromebook hardware moving in some wild new direction that would need Google to light the path. Tensor for Chromebooks will be a sizable inflection point, and when I look at it in this light, it makes total sense for Google to be out of the Pixelbook game until Tensor arrives.
Google is involved in all of it
So, if we can agree that Google’s next real reason to need to manufacture a new, torch-bearing Chromebook is Tensor, the next reason I think they’ll share that same custom SoC revolves around the fact that Google is actually heavily involved in every single Chromebook that exists today. That’s right: whether it’s made by ASUS, Acer, Lenovo or HP, every Chromebook you can buy has Google’s fingerprints all over it. While Chromium OS is the open-source OS that Google maintains, Chrome OS is Google’s version of it and they sign off on every single Chromebook made that comes with it installed.
This is wildly different than what happens with Android phones. Yes, AOSP (Android Open Source Project) is fully open source, and Android with Google Play is a more Google-specific, locked down version, but Google doesn’t get involved with every single device running Android – Google Play included or not. This is quite obvious when you look at the landscape of Android phones and the sheer variation in not only how the OS is handled, but what versions are running on those phones. While it is tempting to think of Chrome OS in the same way, it is not the same situation.
With Chromebooks, Google has to sign off completely on every device and they are deeply involved in the development of many of them along the way. In this way, the chip makers (Intel, MediaTek, AMD and Qualcomm) and the device manufacturers (i.e. – Acer, HP, Samsung, ASUS, Lenovo) truly act as partners with Google when bringing a new Chromebook to market. This isn’t corporate speak, either. The relationship between Google and those that are responsible for building Chromebooks runs deep and is essential for building a wide expanse of Chrome OS hardware that all feels unabashedly Google-y.
Apart from the devices that have gone out of their AUE, all the Chroembooks out there are all on the same version of Chrome OS. All of them will be getting Chrome OS 96 over the next week or so and all of them boot up and operate the same exact software experience. It’s honestly remarkable how well Google has scaled this entire thing to keep so many device up to date and free from manufacturer bloatware and UI tweaks. When you boot up a Chromebook, you know exactly what to expect, and I love that.
Because of this unique and deeply-integrated partnership between Google and hardware manufacturers, the way they utilize their own, in-house Chromebook brand (Pixelbook/Pixel) makes a ton of sense. Google’s not trying to compete with their hardware partners and there’s no real reason for them to start. Google is deeply involved in every Chromebook out there and wants to see them succeed, so why would they constantly introduce products that threaten that success? Answer: they wouldn’t. Instead, I think it is a very good reason to think they might actually share some of that hardware.
What Google sharing Tensor could look like
While the Pixel and Pixelbook lines have always been about lighting the path, I think Tensor could be an opportunity for Google to not only provide inspiration but also some concrete hardware to make these visions reality. After all, if Google’s not arriving on the scene to compete with companies like ASUS, Acer, and Lenovo, wouldn’t all the work to develop Tensor for Chromebooks be a bit of a waste? For the Pixel phones, this totally makes sense because Google is in the phone space to directly compete with companies like Samsung, Apple and OnePlus. With Chromebooks, they simply aren’t. Instead, it benefits them to prop up their manufacturing partners.
So, imagine a scenario where Google launches Tensor for Chromebooks, it is insanely good, and they keep it for themselves in a new Pixelbook. I’d buy one and I bet a bunch of you would, too, but does that reality make any sense after what we just talked about above? Not really. Tensor for Chromebooks would give Google a massive competetive advantage, having a custom SoC not only built specifically for Chromebooks, but built by the company that is responsible for the OS, too. Imagine a spectacular Pixelbook with a super-thin design, crazy battery life, and wildly-good performance and then imagine it being pretty affordable (using your own SoC helps in that category). That would be a competetive device that likely wouldn’t look like a very partner-like move. It would look like direct competition.
I simply don’t think that is Google’s aim at all, and I think they fully plan on keeping things exactly as they are in the Chromebook market. And, if that is the case, it would only make sense that they are spending all the time and effort to make Tensor for Chromebooks in order to share it. Sure, it’ll debut on a new Pixelbook or something like it, but I’d be shocked if Tensor doesn’t become a processor option for other Chromebook makers to utilize once it becomes available.
Imagine for a second a Tensor SoC that is finely-tuned for Chrome OS and imagine how inviting that would be for Chromebook manufacturers. All the work that goes into getting Chrome OS to work with Intel or AMD or Qualcomm or MediaTek simply goes out the door. Tensor is made for this, built from the ground up to run Chrome OS. Ostensibly, developers would only need to get things in place to support their specific hardware and they’d be ready to roll.
In this scenario, would Tensor be the only Chrome OS SoC after some time? I don’t think so, but it’s a way Google could continue building custom chips and prop up the business a bit. After all, they won’t be giving the SoC away for free, so instead of just being a chip maker who keeps things for themselves, they become more of a direct competitor to Intel, MediaTek, AMD and Qualcomm. If Tensor is an option for other manufacturers to use, perhaps it becomes a popular-enough option to almost act as a lighthouse for other chip makers. The message? It is worth your time to build Chrome OS-specific SoCs. MediaTek is already moving that direction and is ahead of the curve. Will Qualcomm, Intel or AMD follow?
Time will tell on that front, but it seems clear to me that Google isn’t building Tensor to keep to itself in the Chromebook space. They are making it to perhaps create a new standard. If it is fast, efficient, affordable and runs Chrome OS better than anything else, chip makers will have no option but to pivot and do the same. A few years ago I would have never thought chip makers would even consider making Chrome OS-optimized SoCs. Now, it looks like that could become the new standard. Either way, we have a bit of time to see this shake out as 2023 feels like a long way away. Things could shift between now and then, but it sure feels like Google is on a path that could really shake up the Chromebook world, and I’m very much looking forward to it.