For retail investors looking to invest in the cloud computing market, they could be forgiven for thinking that it’s an opportunity that has already passed them by. These days, the world’s top cloud infrastructure providers – Amazon Web Services, Microsoft Corp. and Google Cloud – are among the world’s richest publicly traded companies, valued at millions of dollars.
This trio of companies is commonly referred to as the “Big Three” in the public cloud market, and they use their enormous existing cloud infrastructure and their bottomless pit of financial resources to maintain their dominance of it. Should any potential upstart arrive on the scene, such as a novel cloud computing startup, they’ll immediately invest in that company, long before it goes public.
A good example of this can be seen in the AI industry, where startups like OpenAI and Anthropic have emerged as two of the leading lights with powerful AI models like GPT-4o and Claude. In the case of OpenAI, one of its earliest investors was none other than Microsoft, which famously pumped $10 billion into it back in January 2023, shortly after the debut of ChatGPT. Similarly, Anthropic is backed by a number of high-profile venture capital firms, including Google’s VC arm.
Retail investors don’t get to participate in these opportunities. During the early stage of the typical startup’s life, before it goes public and lists its shares on the public stock exchange, the only people who can invest in them are “accredited investors” owned by extremely high net-worth individuals.
However, with the rise of decentralized cloud computing, retail investors may finally have an opportunity to bet on what has long been one of the most lucrative segments of the wider tech industry.
What is Decentralized Cloud Computing?
The integration of blockchain in cloud computing is a logical shift given that these networks live on what is essentially cloud infrastructure. Blockchains are decentralized networks made up of hundreds and sometimes even thousands of independent yet interconnected nodes – basically individual computers and servers – that are scattered around the world. In the case of Bitcoin, for example, it’s estimated that there are around 18,900 active nodes all running the same blockchain, with roughly 1,900 in the U.S., 1,300 in Germany and several hundred in countries like France, Japan, Netherlands, Canada, the UK and Switzerland, according to Bitnodes.io.
Within decentralized networks, the nodes collectively act in the same way as a centralized data center, aggregating their storage to create a single, much larger computing resource. They all work together to perform computing tasks, with a portion of each workload distributed to specific nodes within the network. The advantages of decentralized networks include greater privacy, more reliability and lower costs, though they also face some unique challenges not associated with centralized data centers.
Unlike their centralized counterparts, decentralized cloud networks are built on an alternative business model where users pay to access computing resources using digital tokens. In turn, those who participate in these networks by contributing resources can earn tokens for doing so. These tokens are the key that opens the door to retail investors.
Who Is Building Decentralized Clouds?
One of the leading decentralized cloud computing projects is Cudos Network, which is at the forefront of efforts to create “artificial general intelligence” or AGI systems that can think for themselves and teach themselves new skills.
The creation of AGI will require access to vast amounts of computing power in order to process the enormous datasets required to train their underlying models. Unfortunately, it is not really feasible to meet the computation requirements through centralized cloud providers, due to the high costs involved.
Cudos provides an alternative with its decentralized cloud network, which tokenizes access to both the computing power and the underlying storage resources needed for AGI workloads. The great thing about Cudos is that anyone can join its network by linking their computer to it and donating resources. These resources are then matched with real-world applications that need the computing power they provide. In return, participants earn the platform’s native token, CUDOS, which is a cryptocurrency that’s also required to pay for access to the network’s resources.
Besides AI, any other kind of application can run on the Cudos network, including scientific research, graphic design, machine learning algorithms, games and more.
A similar service to Cudos is provided by Io.net, which was established in 2022 as a decentralized GPU network aimed at supporting AI workloads. Io.net makes use of the same open-source libraries that are used by OpenAI to distribute ChatGPT training across more than 300,000 GPUs and CPUs. It’s built on the Solana blockchain, which is known for having much faster and lower cost transactions than the Ethereum or Bitcoin blockchains. Access to the network is based on a similar model of tokenization, where users must pay in cryptocurrency, with the revenue distributed among network participants.
Doing something slightly different is Render Network, which has built a global rendering system for immersive computing workloads. One of Render’s major customers is Apple Inc., which is using the network to power its Octane X 3D software for creating Hollywood-style CGI and 3D virtual reality worlds. In addition, Render also supports generative AI models for image generation, such as Stability AI’s Stable Diffusion model.
Advantages
The advantages of decentralized networks include lower costs, greater privacy and more reliability, though they also face some unique challenges not associated with centralized data centers. The lower costs stem from the fact that decentralization removes the middleman from the equation. Moreover, overheads such as rent, labor and electricity costs are reduced, as they are spread across the individual network nodes.
According to Ben Goertzel, a renowned computer scientist, AI researcher and the founder and Chief Executive Officer of SingularityNET, there is likely to be lots of demand for decentralized networks in the coming years, as many startups and small developer teams are unable to pay the premiums associated with AWS, Azure and Google Cloud.
“In some ways they offer a great service, but they come at quite the premium if you just look at the hardware firepower you’re getting and not the actual services you’re getting,” Goertzel said. “So we can offer a much more economical solution because of decentralization and also because of the way we’ve set up the software. There’s a lot of upside we can offer to customers.”
By tokenizing access to the cloud, it’s possible for projects like Cudos and Render to distribute cloud resources over a much more varied range of hosts, with a model that’s more accessible and liquid for both customers and those who want to invest.
One of the most revolutionary aspects of data center tokenization is that the compute economy, which has traditionally always been an illiquid asset only available to elite investors, becomes much more accessible and tradable. The ability to invest in this economy is open to every investor, large and small, regardless of their financial resources. Tokenization enables what’s called “fractional ownership”, where computing power can be broken down into more affordable chunks that anyone can afford. A good example of this is DcentAI, which has created a blockchain-based GPU cloud for AI developers, where anyone can buy and sell fractions of its cloud-based resources.
The emergence of the tokenized data center will pave the way for DeFi investors to access a previously restricted segment of the technology markets, namely cloud-based GPUs, which is expected to grow at a compound annual growth rate of 33.2% per year through 2029. In addition, it also provides cloud startups with an alternative funding model to traditional venture capital investments.
Challenges
As promising as data center tokenization sounds, this nascent industry will need to overcome a number of challenges if it’s to become a truly viable option for retail investors. For one thing, the rules and regulations around tokenization are still very unclear, with no standardized legal framework in place to govern digital assets. There’s a distinct lack of clarity over the legal status of digital assets, for example, which creates uncertainty over the safety of such investments.
This lack of regulation is one of the main reasons why the model has yet to gain broad acceptance in financial markets, despite being around in other sectors for a number of years already. For instance, tokenized real estate is not a new concept, yet very few real estate investments are made this way.
An Enticing Opportunity
Still, the potential of tokenization is clear, not only in the data center and cloud computing, but in almost every market. Almost everyone is expecting explosive growth in the AI industry over the next few years, which can help to increase the allure of investing in tokenized data center resources.
Given that retail investors have no other way to participate in the growth of this industry, tokenization represents what may well be the only way for the broader public to invest in the infrastructure that supports our AI-driven future.
