The internet of tomorrow spans Web 3.0, the Internet of Things, 5G mobile networks, VR/AR technology, machine learning, and AI, and other areas that we are already experiencing today such as the Metaverse, Blockchain, and Cryptocurrency.
As the internet becomes more sophisticated, advanced, and widespread, so do the energy consumption requirements needed to run the hardware making it all possible.
This is where it gets tricky.
We keep hearing common claims that the internet of tomorrow will be green. But, anyone with a cursory knowledge of the politics and realities of corporate production and investment in the energy sector can tell you that the needs of tomorrow are nowhere near being halfway met by green resources.
Expecting the internet of tomorrow to naturally gravitate towards “greener” pastures is one thing. If the estimates of risk to the planet and predictions of catastrophic weather disruptions in the near future are accurate, it would be ignorant, and even dangerous, to assume that we will somehow course-correct on autopilot.
It is time to take a dispassionate look at what will be the energy consumption needs of Web 3.0.
What the Heck Is This Web 3.0 Anyway?
The essential feature of Web 3.0 is that it is decentralized. To understand what that really means, let us briefly walk through the history of the internet.
Web 3.0 Evolution: The Static vs Interactive vs Decentralized Internet
The first version of the web, Web 1.0, consisted of static webpages controlled by the creating companies or personages. Today, our Web 2.0 is interactive. Through apps, advanced video and audio real-time input technology, and open-source software, we’ve become more active participants. Instead of just passively reading webpages, we communicate, record, live stream, tweet, blog, publish, hang out and chat. We can collaborate, create, administrate and operate businesses, productions and projects together.
However, content is still mediated by big corporate companies powering the apps which allow all this interaction possible. We need a Twitter to tweet, a Blogger to blog, a Facebook or WhatsApp to connect, a YouTube to live stream, and so on. As such, we are subject to not just control but data mining, and possibly surveillance and manipulation (through ads and other content suggestions on social media).
So how will Web 3.0 be different and how does that affect its energy consumption figures?
The Decentralized Web 3.0: Blockchain Removes Corporate Control and Increases Connectivity
Just like the https and ftps protocols that run the familiar world wide web, the story of Web 3.0 starts with the advent of a different type of protocol: blockchain. The blockchain protocol found a way to use the cross-connections between all the users of the internet, which includes you and me, as the basis to create internet exchanges of information with security.
Until Web 2.0, individual users have to connect to the main company server (such as Google or Meta servers) to get anything done. In Web 3.0, every single individual user should be able to connect directly with other users for all kinds of activities, projects, and collaborations.
This direct connectivity at the level of individual users (called nodes in the internet lingo) instantly has two effects:
- Like a true democracy, it takes the nexus of control from companies only and distributes it across all the users. Much like the concept of shares in a company in a stock exchange. However, anyone with more individual computers (or nodes) connected in the network can accordingly exert more influence on the goings-on.
- It opens the door to widespread reliance on 3D graphics (metaverse and VR/AR apps) and 5G mobile networks and invites more participation in applications involving AI and machine learning.
Where Do the Energy Needs of Web 3.0 Stem From?
Now that we know what constitutes Web 3.0, let us do a breakdown of each feature’s energy demands:
Energy Consumption Needs of Blockchain
Currently, blockchain is most frequently applied in cryptocurrency exchanges online. There are nearly 19000 cryptocurrencies out there and many of the long-heralded ones such as Bitcoin, Ethereum, etc are booming. The total market valuation of this industry is around $1.75 trillion, equivalent to the gross domestic product of Italy, the world’s 8th largest economy.
Know that when we estimate the energy consumption rate of one cryptocurrency, say Bitcoin, we are in essence multiplying the energy needed to secure one transaction involving Bitcoin via the blockchain protocol by how much the volume of transactions Bitcoin is generating (that is how much is it being purchased). According to industry estimates, then, Bitcoin alone consumes between 130 and 150 TetraWatt hours of energy. This usage is higher than how much energy per hour is used in Argentina, a country of 45 million people, and nearly 80% of what the New York State burns up by the hour.
Remember that the Bitcoin market share typically represents 35% of the total cryptocurrency market, so you can roughly triple the TWh to get an estimate of the total energy needs of this industry.
If you wish to find out exactly what it is in Blockchain technology with such huge energy demands, check out my earlier article on the subject with the Cyberpunks, in which I explain in detail the energy-heavy process of creating a single transaction in the blockchain protocol. Thousands of warehouses are set up around the world mining coins and securing exchanges into blockchain ledgers; but also releasing that much CO2 emissions in the air, burning up electricity, and releasing electric waste.
To put things in perspective, a single bitcoin transaction burns 2,292.5 kilowatt hours of electricity. This amount of electricity is what a typical US household will spend over the course of 78 days ( i.e. nearly 2 & a half months!). Note that this is in addition to the carbon footprint left due to the overreliance of the industry on fossil fuels.
Note that noncurrency usages of blockchain are not as high in energy consumption. However, given the upcoming ubiquity of blockchain as the central feature of a decentralized Web 3.0, it would be worth estimating, in a future article, the annual energy consumption of total blockchain usage.
For now, the above Bitcoin figures should leave a rough estimate in the reader’s mind.
Energy Consumption Needs of 5G Mobile Networks
5G mobile networks will be an essential part of Web 3.0. Just like 4G mobiles acclimatized to the demands of an evolving Web 2.0, 5G mobile phones are being designed to accommodate everyday Web 3.0 features. That includes ubiquitous access to 3D graphics (virtual reality), apps relying on complicated machine learning algorithms, and of course expansive data allowances to make these next-level Web 3.0 amenities possible.
Dexter Johnson, an industry insider plainly warns:
โA lurking threat behind the promise of 5G delivering up to 1,000 times as much data as todayโs networks is that 5G could also consume up to 1,000 times as much energy,โ
Dexter Johnson, IEEE Spectrum
The University of Sussex recently published a review of whole-network, energy-consumption assessments of 5G mobile networks. They wanted to see if the evidence actually backed the claims that the industry was aiming to close in on Net Zero (that is a 100% cut-down on industry-generated CO2 emissions) with the 5G mobiles roll-out. The authors found that the evidence backing those claims was sorely lacking. In fact, their conclusions fear that 5G roll-out will only increase the already existing energy consumption burden of mobile phone networks:
“Widespread adoption of unlimited data subscriptions for 5G users and the facilitation of advanced and data-intensive mobile services such as VR and more sophisticated mobile gaming could โencourage energy-intensive user practices, contribute to ever-growing levels of data traffic, and counteract the energy-saving potential of 5G efficiency improvements.”
The University of Sussex review on whole-network, energy consumption assessments of 5G mobile networks.
Several industry insiders have pointed out specific aspects of 5G Mobile Networks with increasing energy consumption needs.
- โA 5G base station is generally expected to consume roughlyย three times as much power as a 4G base station. And more 5G base stations are needed to cover the same area.โย IEEE Spectrum,ย 5Gโs Waveform Is a Battery Vampire
- While the International Telecommunication Union (ITU) has already set up detailed and measurable requirements that qualify a mobile network for 5G status, no comparable standards have been set for energy efficiency.
- 5G networks will usher in an explosion of small cellphones, massive multiple-input-multiple-output (MIMO) antennas to handle the complex data transmission needs, the exponential rise in cloud computing, and mobile interconnectivity with other connected devices and platforms.
- “While there may be improvements in energy efficiency for new devices, these gains are completely lost in the increase in demand for bandwidth, the โmegabytes per secondโ. Consider the huge expected growth in games, videos, streaming services, virtual reality (VR) and the โmetaverseโ” — JRS ECO Wireless Blog.
- According to a study by Vertiv, a US service provider, Telecom providers are estimating an increase in energy consumption needs by 2026 of up to 150-170%.
All this talk of increased energy demands due to high-data-driven applications leads us to the next ballooning energy consumption use case of Web 3.0.
Energy Consumption Needs of 3D Graphics, AI, and Machine Learning
At present, Facebook’s Metaverse is the most well-developed Web 3.0 platform in terms of 3D graphics and deep machine learning AI. It is a virtual reality environment providing various work, business, education, fitness, entertainment, and connection solutions for end-users. It relies on Virtual Reality and Artificial Intelligence, both of which come with costly data processing indicating high energy consumption demands.
According to Statista reports, Facebook’s electricity usage jumped from 3.4 to 5.1 TeraWatt hours from 2018 to 2019 alone. While the company claims to be 100% reliant on renewable energy resources, objective assessments of the type of technology in use behind Facebook’s Metaverse paint a different picture.
Moreover, Intel estimates that the global computing infrastructure will need to be a 1,000 times more powerful than it currently is to actually sustain the metaverse. Does that mean a 1,000 times higher energy consumption needs?
Energy Consumption Needs of a Typical AI Machine Learning Model
A study at the University of Massachusetts “performed a life cycle assessment for training several common large AI models” to compare their carbon footprint. The findings were astounding. The models emitted more than 626,000 pounds of CO2 emissions in total. And that’s five times greater than the lifetime emissions from an average American car.
AI researchers have suspected inexorable energy needs in AI and deep machine learning training for some time. But these findings have surprised other scholars in the field. The models studied in the paper represent a subfield of AI called Natural Language Processing (NLP). All four of these models have been directly involved in some of the huge milestones NLP has surpassed in recent years. However…
“But such advances have required training ever larger models on sprawling data sets of sentences scraped from the internet. The approach is computationally expensiveโand highly energy intensive.”
Karen Hao, Senior AI Editor, MIT Technology Review
Energy Consumption Solutions Lag Far Behind the Speed of Web 3.0 Evolution
Three things have stood out in the course of my research for this overview:
- Tracking the energy consumption figures of the hardware running Web 3.0 and its associated facilities is not a widespread practice.
- Even industry insiders express shock and concern when evidence or estimates of actual energy consumption of some of the planned features of Web 3.0 and 5G are reported.
- Current evidence of the IT and mobile energy going for renewable energy resources which are less climate-damaging is sorely lacking and always falls short of the claims.
- The gap between the ballooning energy consumption needs of tomorrow’s internet and the effort to replace high energy-consuming hardware with more efficient hardware that runs on green energy is growing wider.
In sum, the pace of software development is lightspeed compared to the pace of the hardware catching up in its energy efficiency, or the industry catching up in its resolve to actually switch completely to 100% green resources (compared to traditional fossil fuels).
We need to turn the talk of energy consumption of the internet’s future into a trend. It’s a metric that can help us decide what’s a massive energy drain all things considered and what’s less so with more real-world real-people utility.
This column will certainly keep returning to this topic in the future.
That’s true, after the emergence of the latest technologies, people who are getting into software development actually stay ahead, but I can say that it’s actually a challenge to find decent specialists in that field. It took me a while to discover Sloboda Studio, but at least it was all worth it, so I’m pretty sure that the number of qualified specialists will increase a lot.
I agree that the number of reliable developers in that field is definitely increasing, but I can say that even nowadays, it’s not a problem to find good specialists in that field and implement top-tier technologies to your business processes. In my case, it wasn’t a problem to discover custom application software development company Uran, and they provided me with a great business solution for my company.
I don’t really understand how Web3 works, but I do know that projects like wakweli.com use blockchain technology and Web3 to create an investor solution, a standard of trust, and certification of unique tokens. I really believe that such solutions will make it easier to invest, as there will be no doubt about fraud.
IoT predictive maintenance is revolutionizing the way organizations approach equipment maintenance, enabling proactive measures and data-driven decision making. While organizations face challenges in implementing and managing IoT predictive maintenance systems, IoT service providers play a vital role in providing end-to-end solutions and support. Their expertise, hardware offerings, analytics capabilities, and ongoing maintenance and support services contribute to the success of predictive maintenance initiatives. By collaborating with IoT service providers, organizations can leverage the power of IoT to enhance efficiency, reduce downtime, and optimize their maintenance processes, ultimately leading to increased operational excellence and competitiveness in today’s evolving business landscape. Read more – https://a-team.global/blog/iot-predictive-maintenance/ here.
I see how you highlight the evolution of the web from static to interactive, 2.0 to 3.0, emphasizing the control exerted by big corporate companies over the content and functionality of these interactive platforms. This control can include data mining, surveillance, and manipulation through targeted advertising and content suggestions. Interesting topic. It was 2.0 and then transformed to 3.0 web,
I like to see concepts based on real examples. As I understand, food delivery app development primarily falls within the scope of Web 2.0, where users actively participate, communicate, and interact with various platforms and services. It enables users to order food, track deliveries, make payments, and provide feedback, among other interactive features. I think technology will continue to evolve to Web 3.0, which shows decentralization and user control. There could be potential implications for food delivery apps and other web-based services. As I understand, Web 3.0 aims to provide more transparency, user privacy, and data ownership by leveraging technologies like blockchain and decentralized networks.
Great article! It’s insightful to see the environmental impact of Web 3.0. Balancing innovation with sustainability is crucial, and understanding energy consumption is a step in the right direction. Looking forward to more discussions on making blockchain technology greener.