I came across RNDR token in late 2020 before the enormous NFT swell. I recognised almost immediately the enormous potential for this bad lad. It stood out amongst a lot of other tokens, because it actually solves a genuine problem, namely: how do you supply the world with rendering power when tech companies don’t want to do it?
Most tokens are fairly straight-forward to understand. Some are even direct de-centralised knock-offs. It does not take Albert Einstein to know that Filecoin is a de-centralised Dropbox, or that Theta is a de-centralised Youtube. That’s what makes RNDR pretty special. There is some homework required to learn about the problem RNDR is solving. You need to delve a little into 3D graphics, the 3D animation pipeline and the overall the 3D eco-system. Once you do, it will become clear that this RNDR thing something very special.
So, lets jump right into it knights…
3D, Rendering And RNDR’s Role
3D graphics are three-dimensional representations of geometric data. 3D graphics were originally the focus for animators, pioneered in part by George Lucas effects company “Industrial Light & Magic. The biggest use for 3D graphics today is of course the gaming industry, which is dominated by gaming giants such as Epic and Unity. These company’s gaming engines are now the most powerful 3D engines in the world.
The most important part of the 3D graphics creation pipeline is a process known as rendering. Once a 3D graphics model has been created, the rendering process is needed to make the 3D graphic appear real. During the rendering process, the graphics processing unit of a computer (GPU) assists the CPU in performing complex rendering calculations. The end result is a photo-realistic graphic.
Still awake you festering medieval vassal?
Alas, I continue.
Beyonce has made it clear throughout the years that girls run the world. But who runs global rendering power? Big tech companies like Apple, Facebook and Google will simply build data centres filled with GPUs, right???
Eh, yeah. Sort of.
Cloud giants like google and amazon supply the world with GPU data centres… but they won’t forever.
The GPU Problem
Nvidia and AMD are also big players in the 3D world - they control most of the GPU market. The powers that be have collectively decided that enterprise GPU computing should be significantly more expensive than consumer grade GPU power.
I am guessing that this is a way to squeeze the big boys for pennies…
Big tech companies have large coffers - why not take them to the cleaners?
Lets take a quick look at two random GPUs released within a year of each other: the consumer-grade RTX 2060 (released January 2019) and the enterprise-grade T4 GPU (released in September 2018).
According to versus.com, the RTX 2060 crushes the T4 on many counts.
But in addition to these factors, it is also significantly less. At today’s prices, the RTX 2060 costs $567 USD, whereas the T4 costs $2,711 USD.
TL;DR here is: consumers are getting more powerful, cheaper GPUs than tech companies…
The net effect of this significant pricing difference is that tech CEOs do not want to build large data centres of GPUs. You spend enormous sums of money on GPUs, the technology basically expires after two years - why bother?
Aside from the enormous capital cost of buying data-centres of GPUs, the cost of electricity is very significant. GPU requires a serious amount of electricity to power the units.
These economics have not gone unnoticed by OTOY founder and mad scientist Jules Urbach. Urbach spent years building GPU data-centres for tech companies, culminating in a collaboration with Amazon in 2013 with their EC2. Urbach noted the lack of enthusiasm amongst tech companies for building GPU data centres and the surging need for rendering power. His solution was simple:
to link the cheaper, consumer grade GPUs with rendering jobs around the world - linking both parties via the RNDR blockchain network.
The RNDR Network
The RNDR network is a simple idea that takes the “proof of work” concept from bitcoin and overlays it onto a rendering service. “Proof of work” is a monstrous mis-use of energy resources that involves crypto “miners” solving pointless puzzles in order to be rewarded blocks. OTOY’s “proof of render” framework allows for someone to enter a rendering job, approve it and pay the renderer in RNDR tokens.There is no waste of energy: people pay to render files and the miners get dollar-fucking-bills yo.
The RNDR network has almost no margins too - it only charges a .5% network fee. Also like other blockchain projects, it will only be profitable to mine in jurisdictions with very very cheap electricity. If you combine a zero margin business running from the cheapest electricity regions in the world, you get a cloud rendering service which virtually no cloud tech company could plausibly compete with. AWS runs on a 30% margin. Will they will be bothered running a cloud rendering business with these margins while competing against RNDR? Eh, no:
Good with getting finance to play ball with you on that one. And don’t expect Sheila from legal to be so cutesy with you anymore in the future either.
Yet I regress.
Instead of competing with RNDR, many of the big tech companies are partnering with them to use their services.
The Coming Surge In Rendering Demand
For those outside the warm nexus of 3D-focused companies such as NVIDA, Epic Games, AMD and Unity… you might think:
Rendering… 3D… Why should I give a fiddler’s elbow about all this stuff???
Well because we are on the forefront of a 3D revolution.
Come the day and come the hour… we will very much needeth rendering power to power this stuff.
Now gaming is the most obvious source of rendering demand - all game developers need to render their 3D models before introducing them into games. The gaming market is also no mean feat - it is a 173 billion dollar juggernaut that is growing fast.
The demand for rendering power is mot limited to gaming though - it is coming from the convergence of the gaming world with literally everything. As the a16z guys might say: “gaming is eating the world”.
Virtual Production is one such example. Virtual production is the integration of gaming engines into the move-making process. 3D graphics have become so good that studios can now take 3D images from gaming engines and project them onto LED walls to create film sets. This cuts the cost of filming significantly and means film crews no longer need to film on actual sets.
As graphics become better, it will eventually become standard fare for studios to run all their special effects through 3D gaming engines like Unreal Engine - again cutting costs dramatically.
To demonstrate the raw power of gaming engines in the film-making process, lets look at a project in detail. Gods of Mars is a feature film directed by Peter Hyoguchi and Emmy Award-winner Jonathan Schriber that is currently in (virtual) production. The strategy for the film is to create it entirely within Unreal Engine, leaning on the VFX and 3D suite that brought blockbuster games like Fortnite. Other projects such as The Mandalorian have used Unreal previously to avoid filming in different ‘locations’by projecting them onto LED walls. Gods of Mars goes further by using Unreal for almost everything basically.
Our good friend the director Hyoguchi is not doing this for the good of his health. He is doing this to save a fuck load of cash. Hyoguchi explains in detail below.
And no, I don’t understand the first sentence either. Fucking deal with it:
For production, zero company moves provide incredible savings — travel, insurance, permits, housing, etc. No need for building full sets also cuts massive spending from the budget. Not shooting with green screens cuts out massive post-production VFX compositing fees.
All told, this new visual reality approach is running only about 30% of the costs of a typical CGI heavy production. That is a huge potential saving on a project that normally could cost north of $200 million.
I don’t know about all the other stuff, but my ears perked up when I heard “30% of the costs”
The opportunities for RNDR in film-making extend beyond special effects. The introduction of avatars within the film-making process would allow aging/deceased actors to be introduced into films. This means it is not the end of the road for James Dean, or Robin Williams: they can re-introduced into the virtual product pipeline using Unreal Engine. While Unreal does not currently offer this set of tools, it’s Meta Human app is an early indication of the direction Epic are going: avatars are a big part of their plan. OTOY themselves are excited about these prospects as they dubbed virtual production digital avatars “the holy grail of game/film technology crossovers.”
Spatial media is another burgeoning opportunity for RNDR. Companies and individual creators will need to lean heavily on 3D gaming engines to produce virtual and reality content. 3D gaming engines will continue to ‘eat the world’ as knowledge of Unreal/Unity becomes a prerequisite for any form of creation. It is not impossible to imagine a future where gaming engine knowledge is a basic requirement for anyone looking to create. Could knowledge of Unreal Engine be the equivalent of understand Adobe Creative Suite in 2010s? Every 3D model used in augmented reality and virtual reality will need to be rendered and RNDR will be offer the fastest/cheapest form of rendering power. This could easily end up as a winner takes all type beat.
The most esoteric and exciting and over-hyped use case for the surge in rendering power is to power the “metaverse”. Many have fallen in love with Tim Sweeney’s vision (expressed by Matthew Ball) for the Open Metaverse underpinned by blockchain technology. For those that believe this vision is achievable, RNDR token will be absolutely essential. Every digital good within these virtual world will need to be rendered and again RNDR will offer the cheapest form of rendering power via a useable client. RNDR’s moat is not limited to the power however. RNDR will also allow creators to ‘watermark’ their work to prevent copycats from reproducing their work. We see the importance of copyright protection in Second Life, where copying has run rampant in the past.
In 2006, Second Life creators noticed a strange bug within the system - a copybot which allowed all 3D items to be copied. This meant that users could basically copy any 3D model and sell it at a profit. Authenticity is an incredibly important component of a virtual economy that largely consists of selling virtual goods. The copybot therefore wreaked havoc on Linden Lab’s virtual economy, because every item could be re-produced for free. It is not difficult to see why this fucked shit up. What would the world look like if you could copy every item within a department store, or shop? A country’s commercial structure would quickly descend into anarchy. While Linden Labs eventually managed to shut down the copybot, they were not able to solve the issue of watermarking. According to Making of Second Life author James Wagner A.U, they tried creative commons licensing and creation date watermarks to bring an end to the disruption in one year. Nothing worked.
The moral of the story is your mother. But the real moral of the story is that RNDR token’s water marks solve a problem that is hidden for now, but it is likely to grow in importance as the de-centralised metaverse emerges, namely the issue of copycatting and fraud. We saw early indications of this during the 2021 NFT boom, where a fake banksy has netted over $1 million in ETH!
If you are an NFT/virtual goods creator, you will therefore need to RNDR. First for cheap rendering power, then to stop copy cats from flogging your shit by using their watermark system.
There is much that I have excluded in my RNDR write up. It is worth mentioning that OTOY founder Jules Urbach is perhaps the most qualified person on the planet to build a decentralised GPU work. Not least because he actually came up with the idea for a decentralised GPU network in 2010 - a year and a half after bitcoin was created. His 2010 patent described “methods, apparatuses and systems directed to hosting, on a computer system, a plurality of application instances, each application instance corresponding to a remote client application”. This computer system is basically the Ethereum virtual machine, which RNDR now runs on as a ERC-20 token. The word “visionary” does not come close.
It is not all rosy gardens though. RNDR could face meaningful competition from Polish titans Golem. Golem aims to distribute the world’s computing power and began by focusing on rendering power - something the team amusingly considered ‘low hanging fruit’ in their company blog. Another big question is is whether ERC-20 tokens are good investments. It seems that investing crypto shares more with religion/tribalism than it does with prudent investing. If you wade in, be aware that are channelling your inner wallstreetbets and not your inner Warren Buffett.
But you already knew.
Or at least I hope you did.
RNDR remains the most exciting crypto project I have come across. It is providing much-needed rendering power at a price that few tech companies can meaningfully compete with. Its Octane rendering client and its deep network of creators/executives/investors in the US give it a moat against competition from the crypto world. Additionally, projects like Golem would struggle to keep with OTOY’s experience and network within the rendering domain.
While the slightly complicated science behind CGI rendering might put off a few folks, however I think it adds to the project. This gives me ample opportunities to accumulate a RNDR warchest while the rest of the crowd YOLOs Cardano.
Disclaimer: I do not currently own RNDR. But I very much plan to be absolutely up to my tits in it… in the coming months and years.