Almost Reality

At CCS Insight we’ve been reassessing our global forecast for virtual reality (VR) and augmented reality (AR) devices, and our latest update will land in the next few weeks. As part of that process, we’ve been debating the timing of AR smart glasses. When will these devices really hit the mass market and go from the “next big thing” in tech to a mainstream consumer offering?

Picking an exact date remains difficult, but there have been strong signs recently that pieces of the very complex puzzle required for AR technology are slotting into place.

For example, in Samsung’s third-quarter results, the company mentioned that the VR and AR market “is projected to enjoy full-fledged growth starting in 2023”, a surprisingly direct statement from a company that typically keeps its cards close to its chest. Another noteworthy development came from Vuzix, which makes smart glasses and waveguides — ready-made displays for AR devices. It intends to expand manufacturing of these displays following increased customer demand.

Although these hints are useful, I feel the most important indicator we’ve seen recently came at the Snapdragon Summit in November, where Qualcomm lifted the lid on the Snapdragon AR2 Gen 1 chipset, the company’s first offering built specifically for AR glasses.

This silicon design is built for distributed processing; this means that computing is split between the glasses and a host device such as a smartphone or laptop, connected over Wi-Fi. The chipset design is made up of three separate components:

  • The main AR processor, which manages tasks such as user tracking and environment sensing thanks to a 4 nm process node
  • A co-processor, which handles sensors such as cameras, as well as artificial intelligence capabilities
  • A Qualcomm FastConnect 7800 module, which includes Wi-Fi 7, for connecting to a host device.

This approach neatly solves several of the problems that have plagued single-chip solutions. The split componentry allows for a smaller circuit board and reduces the number of wires running around the frame of the glasses. These features make it easier to design and engineer smaller and sleeker devices. Furthermore, wireless split processing means that a greater share of the computing power can be hosted outside the headset, reducing challenges with thermal dissipation.

There’s already a strong list of device-makers signed up to build with this chipset. This includes established players in the VR and AR space such as Nreal, Pico and Vuzix, as well as broader cross-category players like Lenovo, LG, Oppo and Xiaomi. Niantic has already shown a reference design for an AR headset, demonstrating its use for outdoor gaming experiences.

I’m not convinced gaming will be a critical use for AR, but the reference design shows what the technology can do and indicates Niantic is the furthest along in headset development. When it comes to more-mature prototypes and commercial devices, we don’t have exact timescales just yet, but 2023 seems like a reasonable window for these companies to aim for.

Added to this, Qualcomm has been laying the groundwork for a wider ecosystem for AR. The company’s Snapdragon Spaces platform and software development kit have been open for some time and have incentivized developers to build AR experiences that should gel seamlessly with these forthcoming devices.

Put all this together, we have an effort from Qualcomm and its partners to work together on an early play for the consumer AR market, bidding to make a success of it before other major consumer electronics players — most obviously Apple — enter the fray.

We’ve heard Qualcomm and others argue for some time that VR and AR devices will provide the next generation of mobile computing solutions. Many agree that AR is more important on this front because it’s preferable for use on the go as opposed to VR, which is better in fixed locations. As such, any developments in AR glasses are well worth watching out for.

Undoubtedly, there’s need for realism here. We still don’t know what the battery life of these devices will be, and I’m concerned that use might be limited to minutes rather than hours. Furthermore, there’s the question of cost. Unless these glasses can reach a point where they’re seen as an affordable add-on to a host device like a smartphone, they’re likely to struggle — especially in the current macroeconomic climate.

However, I’d argue that anyone who has tried recent VR or AR devices can see the potential here. Even if devices aren’t perfect yet, hardware and software are both moving in the right direction and offering increasingly impressive experiences, as I explored in my recent hands-on time with the Meta Quest Pro. Qualcomm’s investment in the full stack with both its silicon and developer platform bodes well, and it really does begin to seem that the many challenges with AR glasses are being ticked off one by one. It feels like AR is on track to become a reality.