Skip to main content

In 2020, Qualcomm’s slower chips may be more important than its best ones

In 2020, Qualcomm’s slower chips may be more important than its best ones


Here’s what’s next for Android phone processors

Share this story

Qualcomm’s annual Hawaiian press event to announce its products for the coming year is happening. It‘s there in part because Qualcomm needs something equally convenient for Asian and American attendees and also because the company probably is still smarting from the reaction to its absolutely bonkers attempt to make CES its big annual keynote in 2013 (if you don’t know what I am referring to, you absolutely should click).

As I briefly mentioned yesterday, the main thing that strikes me this year is how much more real it seems. Qualcomm is like Intel in one regard: it is several steps removed from being able to directly deliver consumer products, so it has to resort to big promises it can’t actually directly make happen.

Last year’s event was a perfect example: the 5G demos on site turned out to be a big dud, my colleague Sean Hollister reported. A year later and we have real 5G rollouts in the US and actual 5G phones for sale. I would have put even odds that neither would have happened. Much of it is still more hype than reality, but the reality is there. Verizon’s coverage maps show that it exists on a few street corners in major cities. Which, technically, is real 5G if you can find it. Ahem.

T-Mobile really has rolled 5G out across much of the US with its more expansive spectrum, but there are precious few phones that support it — and we won’t really know whether it’s significantly different from LTE until more phones do.

That’s where Qualcomm comes in. It is promising that the Snapdragon 865 will bring 5G, camera, and gaming improvements. That processor doesn’t have an integrated modem at all, and the conventional wisdom is that everybody who makes Android phones will just opt to use Qualcomm’s new 5G modem with it. iPhones will also have 5G “as fast as we can,” according to Qualcomm’s CEO.

Which means that next year we should have broader releases of phones that support 5G, plus networks that actually provide it. 5G is happening, but despite years of buildup and hype I still couldn’t tell you what it’s actually going to mean to the average consumer. And I still believe anybody who says they can tell you is likely trying to sell you something. It’s 5G, they’re trying to sell you 5G.

Beyond working with Qualcomm’s new 5G modem, the 865 is important for purely Android-focused technical reasons. (There is a long digression about how Intel was pushed out of the 5G modem business and had to sell to Apple that goes here, but that’s a rant for another time.) Chaim Gartenberg laid out some of the key details:

As expected, there’s a new ISP (image signal processor), the Spectra 480. Qualcomm’s big spec here is that the Spectra 480 supports “2 gigapixels per second” speeds, which it says enables a host of new photography features.
Phones with a Snapdragon 865 will be able to shoot 200-megapixel photos, capture 8K video, and shoot 960 fps slow-motion video at 720p resolution. Additionally, the new processor will support video capture with Dolby Vision HDR, a first for mobile devices.
Of course, all that requires phone manufacturers to actually meet Qualcomm with the camera hardware to shoot those kinds of pictures and videos, but the Snapdragon 865 at least lays the groundwork by supporting these features right out of the box.

So that’s the chip for every flagship Android phone, one I’m sure will be pretty good and I’m double sure will pale in comparison to the capabilities of whatever A-series processor Apple puts out for its iOS devices next year. So it has always been and so it will continue be.

The fastest processor with the most features isn’t necessarily the best one

I think that Qualcomm’s efforts to elevate the midrange Android chips are much more exciting than what it’s doing on its flagship chip. So I’m going to be paying much more attention to the Snapdragon 765 chip, which could usher in the first affordable 5G phones.

I know this is naive, but here I go anyway: I hope 2020 is the year when people stop assuming the fastest processor with the most features is the best one. There are different chips for different situations. We don’t argue over screen sizes, pixel density, or megapixels anymore. We understand that the bigger number doesn’t directly translate to the better experience.

In theory, anyway. In practice, some of these midrange processors end up being real dogs. I’m optimistic this one won’t be, but we’ll obviously have to wait to see.

Maybe the upcoming Motorola Razr will help turn us away from the overemphasis on phone processor speeds. It’s a high-end device, no doubt, but it has a mid-range chip inside it. The Razr has a clear and well-justified reason for using a slower chip: thinness and battery life. If the actual experience of the phone is good, I don’t think that chip will be a knock against it. We shouldn’t expect the Razr to outperform a 6-inch-plus Android slabphone, just like you wouldn’t expect an ultralight laptop to outperform a gaming laptop.

Trading raw horsepower almost nobody uses for battery life is a deal I think a lot of people will be willing to make.

In fact, it’s possible that the Snapdragon 765 will enable better 5G experiences than phones with the 865. That’s because, unlike the Snapdragon 865, the 765 has a less powerful X52 modem. It’s capable of lower speeds (maxing out at 3.7 Gbps, instead of the 7 Gbps the X55 is theoretically capable of). But it has a big advantage: that 5G modem is integrated directly in the 765 chipset, meaning it should offer improved power efficiency (and, therefore, battery life) than the X55, which is its own separate chip.

We’ve been through this cycle of trying to minimize the focus on processor cycles (sorry) before. There’s even a catchphrase for it: the megahertz myth. That PC-centric context doesn’t strictly apply to phones, but the general idea is the same.

Qualcomm and carriers and phone makers are going to be excited to push 5G harder than ever in the next year because they will actually have real 5G things to sell you. Bully for them. For you, though, the thing to be excited about is that more affordable phones are going to be better than they’ve ever been.

More from The Verge

+ Microsoft planning second next-gen Xbox that’s cheaper and less powerful

My personal preference is an option to buy the top-spec device without a spinning disc drive, but I get the strategy here and would probably do the same thing if I were in Microsoft’s place. I fully recognize that this is contradictory to everything I just said about phone specs. I am large, I contain multitudes.

Here’s a prediction, though: a lot of people are going to do what I am going to do — remember how badly they were burned by buying the Xbox One on launch day. It was so big and dopey and VCR-like, but ultimately what killed it is the lack of games. I am not suggesting any of those things will be remotely true of the upcoming Xboxes. I actually think Microsoft learned those lessons and will do whatever it can to not repeat those mistakes.

But that memory is going to make a lot of people take a wait-and-see approach to the next generation of Xboxes, possibly to Sony’s benefit. If there are any stops left sitting around Xbox headquarters, Microsoft needs to pull them out for this launch. It needs to do more than craft a redemption/comeback story, it needs to make a huge splash that overwhelms entirely justified qualms.

+ There’s no quick fix to find racial bias in health care algorithms

Please welcome Nicole Wetsman to The Verge with a warm hello on Twitter. Her first story for us is a good one:

Bias identification tools can help to make algorithms and artificial intelligence tools more accurate, but they don’t necessarily tackle the root causes of bias, which are larger systemic issues and inequities. An open-source tool might look for problems with an algorithm, but it would not be able to identify issues with the way it was constructed

+ Elon Musk tries to explain Twitter in ‘pedo guy’ defamation case

Liz Lopatto’s first missive from the Elon Musk trial came in just after I put yesterday’s newsletter to bed. As I read it, my face was stuck in a permanent state of disbelieving-guy-dot-gif. It helped immensely that Liz’s writing here is so good.

+ The rise, disappearance, and retirement of Google co-founders Larry Page and Sergey Brin

Nick Statt with a fairly comprehensive set of notable moments and weird anecdotes from the co-founders’ tenure.

+ Larry Page is the tech world’s Dr. Manhattan

If you don’t subscribe to our other daily newsletter, The Interface with Casey Newton, you missed out on this hilariously accurate extended metaphor. Casey also pointed out something I should have:

One popular theory was that they are fleeing — something. Some speculated that they are fleeing a board investigation into Google’s troubling history of inappropriate relationships between Google’s mostly male C-suite and their subordinates. (If you don’t know the name Amanda Rosenberg, you should.)

+ The dark side of electronic waste recycling

You want to read this story by Colin Lecher. It’s about e-waste but it’s also a caper using GPS trackers to uncover corruption amongst friends and the trial that resulted. Great video from Verge Science, too:

“It was very disappointing,” he tells me. Total Reclaim wasn’t just an example of a company seemingly doing everything right. It was run by friends. “Probably one of the most troubling things I’ve experienced in this business of being an advocate was getting a real ally,” he says, “and to find out that you were betrayed.”

+ This robot lets you remotely surveil your cat and track its daily step count

If this thing also launched cat treats like a pellet gun, it would be a complete package.

+ BMW reverses decision to charge vehicle owners an annual fee to use Apple CarPlay

This was up there among the stupidest subscription fees in tech.