Contact Sales
Let us know who you are...
...and we'll reach out to understand your needs.
This post was also featured on Substack.
Next month, it will be a year since Facebook announced the biggest rebrand in the history of, well, branding. No company of such size, stature, and cultural reach has ever made such a bold move, with such a specific strategic purpose — to re-orient the company toward widespread adoption of an immersive digital world known as the Metaverse.
The “new” company, Meta, is not an outlier. Over $120B in investment has been made into the Metaverse just in 2022 so far, dwarfing Meta’s $10B annual spend on its Facebook Reality Labs division.
Clearly there’s smart money betting big here —and some analysis suggests the space could generate $5T (that’s trillion) in value by 2030.
Importantly, there’s a lot of cool stuff already happening in the space — such a massive investment hype cycle without validation from working products and services would be concerning — BUT… no one can disagree that the vision is way ahead of the current state. We’re a long way from the Metaverse. Turning these hundreds of billions of investment into trillions in value is far from a done deal.
Achieving the Metaverse is a complex endeavor that requires a brand new stack of innovations that are diverse not only technologically (major breakthroughs in hardware, software, and networking) but also span a range of types of companies that can deliver them. Interoperability and regulation are huge question marks. Thousands of ingredients must be delivered by hundreds of players, and we’re figuring out the recipe as we go.
There are many layers in the metaverse tech stack.
The last thing any innovator wants to do is to fail slow, and the last thing the investment community wants is to dump hundreds of billions into a complex long-term project where not only do we not reap benefits until everything in the recipe is done, but we can’t even prove that markets support this level of investment.
I. On its current trajectory, the “project” of building the Metaverse is going to be long, risky, and very expensive.
Taking a hard look in the mirror — despite all the excitement and investment, there are many elements of the pre-Metaverse landscape that fit some difficult and discouraging patterns. There are chicken-and-egg problems at every interface in the stack. So how do we keep the metaverse from becoming a decades-long money pit?
The answer is strikingly simple —
Focus on delivering an awesome experience of being in the metaverse.
II. Metaverse investment is significantly de-risked by focusing on the actual experience at point of use.
Iterate on that experience as much as is necessary to achieve “product-market fit”, or perhaps (humor me) “experience-user fit”. This experience leads to user adoption in terms of any relevant metric, even if not all the pieces are completely in place. The rest of the Metaverse tech stack will crystallize around this seed.
So what should that experience look like, and feel like? What does it mean to focus on delivering it?
First, let’s agree on what the metaverse is. Matthew Ball’s definition is
“…a massively scaled and interoperable network of real-time rendered 3D virtual worlds which can be experienced synchronously and persistently by an effectively unlimited number of users with an individual sense of presence, and with continuity of data, such as identity, history, entitlements, objects, communications, and payments.”
I agree with this definition.
Ball continues:
“Most commonly, the Metaverse is mis-described as virtual reality. In truth, virtual reality is merely a way to experience the Metaverse.”
Here, I disagree. “Real-time rendered 3D virtual worlds” are far less compelling when not experienced in actual 3D. And what good is “individual sense of presence” without the feeling of actual presence that can only come from the immersion that VR provides?
III. A critical mass of Metaverse adoption can only be achieved through immersion — through VR.
The kinds of user numbers and user hours necessary to bridge us from the Innovators part of the Tech Adoption Curve, and deep into the Early Majority, will demand an immersive, 3D experience. The 100 millionth citizen of the Metaverse will need to forget, for hours every day, that they are in a virtual world.
I might go so far as to say: those who wish us to believe that a non-VR Metaverse is a viable vehicle for driving us up the Adoption Curve (and for delivering killer ROI on today’s Metaverse investments) are the same people that have a vested interest in us believing that to be true — those who are already invested in keeping the pre-Metaverse gravy train on the tracks. Caveat emptor.
So again, what should our optimized Metaverse experience look like, and feel like, now that we’ve decided it must be an immersive 3D VR experience?
This might seem like a simple question with a simple answer: an immersive experience is the suspension of reality via the perception of being physically present in a non-physical world.
All of the big money and big companies in the space know this. All of the big money and big companies in the space want this.
And yet the most obvious public face of the Metaverse is busy getting kicked in the teeth over showcasing a most assuredly not immersive, low-fidelity VR experience.
Do we really think this space can grow on the back of cartoonish low-fidelity graphics?
Have we forgotten the “reality” part of Virtual Reality? Why the near-total failure to deliver on this promise?
Why do Metaverse graphics still suck?
In short, the real-time VR rendering required for immersive interactivity in the Metaverse is up against laws-of-physics hardware constraints that will not ease soon.
IV. On its current trajectory, immersive real-time interactive VR will fail to even approach expectations due to fundamental hardware constraints.
Let’s break this down:
Therefore: we’re left with extremely low-fidelity real-time graphics on board consumer-grade standalone headsets (enterprise headsets are a bit better, but still not immersion-grade — and are much more expensive).
What is to be done? Let’s start by looking at our options.
Drastically simplifying VR here — but to set a simple baseline, there are 4 primary components that comprise a VR experience:
In today’s traditional standalone VR model — e.g. Meta’s Quest 2 — all these components are handled by the headset.
Traditional Standalone VR:
Again, the Graphics function in particular is hobbled by the seriously underpowered GPU in a standalone headset, with no hope of that changing any time soon. The hardware just can’t handle immersion-level graphics in real time.
V. Traditional Standalone VR cannot and will not handle immersive real-time interactive experiences.
But does the GPU have to be in the headset?
Another approach has been to offload all the computing — both the Application and the Graphics — to a separate PC, with the headset communicating with that PC through a hard wire (e.g. Oculus Link) or over wifi (e.g. Oculus Air Link).
Offloaded PC VR (Link / Air Link):
Being wired to a PC was never ideal — no one wants to do an awesome 360 move in VR only to be strangled IRL — but Air Link seems to work pretty well. In theory, a PC with a good discrete GPU should have the horsepower to handle real-time interactive metaverse graphics. So, problem solved?
Definitely not. Because the application has to be developed for, installed on, and run on an entirely separate device and platform. And that platform, PC VR, is all but dead.
So for the “Offloaded PC VR” approach to work, developers will have to:
And, users will have to:
VI. Offloaded PC VR is a dying branch in Metaverse evolution, and is clearly not the future of VR.
But what if we could:
What would be ideal is if we could offload just the graphics, leaving the application running on the standalone headset…
Remote Virtual GPU:
…and if we could offload our graphics workload not only to a local PC, but alternatively to any IP-addressable device anywhere that has a powerful GPU — it could be a game console, or a GPU shared at the edge or from the cloud.
This might sound like an extreme take, but there is no other viable approach on the horizon.
VII. Remote Virtual GPU is the only credible approach to support immersive real-time VR experiences for mainstream markets — and thus the only way forward for the Metaverse.
Even better, this approach is future-proofed as the experience of VR evolves to include more and more AI (which also requires GPU horsepower) —
Want to do real-time upsampling to enrich the immersion? More GPU is better. Want to mitigate the no-legs-in-VR problem by having the machine assume what your pose might be? More GPU is better.
VIII. Remote Virtual GPU is the only credible way to enrich VR immersion with AI — and thus the only way forward for the Metaverse.
No matter where the immersive 3D Metaverse is headed, more GPU capacity will always make for a better experience.
The sooner we design our Metaverse stack so that graphical workloads are handled outside of the standalone headset, the better chance we deliver on the promise of the Metaverse — before the hype cycle goes bust.
[OK, but how? Our team at Juice Labs is solving remote virtual GPU (a.k.a. “GPU over IP”) once and for all, with just easy-to-install software.]
