<– Network      Go to ToC     Governance–>

Energy is what enables Devices and Networks to perform their irreplaceable role in the development of the Metaverse. The energy aspect is considered from the viewpoint of the ability to:

  1. Store energy for use by mobile/portable devices.
  2. Enable energy consumption by:
    1. An IT equipment to process Data
    2. A network to carry bits.
Storage
Consumption

1      Storage

Storage for portable devices is important to enable portable device with a long time between battery recharges. Today’s predominant battery technology for electronic devices is Lithium-ion operating as follows:

  1. When a Lithium-ion battery is charged, Lithium ions move from the cathode’s Lithium Cobalt Oxide (LiCO2) to the anode’s carbon.
  2. When the battery provides energy, the ions move back to the cathode.

In early Li-ion batteries, a liquid electrolyte was used to separated anode and cathode. In later batteries, a porous separator soaked in an electrolytic gel allowed thinner batteries. The separator of later Lithium polymer cells used a solid polymer but with longer charging times.

While the preferred power storage technology for consumer electronics devices, Lithium-ion batteries have limitations:

  1. Battery recharging can be done a limited number of times because some ions remain captured in the material. The battery fails to operate when free ions are no more available.
  2. Battery charging enlarges the anode’s volume leading to a decrease of the battery’s ability to store ions.
  3. Energy density improvement is slow, as seen from the case of the iPhone: the iPhone1 had a 5.18Wh battery and the iPhone 6S’s has a 6.55Wh battery. It took 8 years to get an improvement of just ~ 26%.

Alternative materials are used for positive and negative electrodes, and electrolytes to obtain denser packing of Lithium ions in electrodes, facilitate their movement through the electrolyte, and improve safety.

For the future, different approaches are studied:

  1. Lithium-Sulphur (Li-S) batteries improve today’s Li-ion batteries thanks to 3D electrodes and nanomaterials with larger energy density. The theoretical energy density limit of Li-ion batteries is ~ 320 Wh/kg against the current ~ 200 Wh/kg, but the theoretical limit for Li-S is ~ 500 Wh/kg because Sulphur can hold two Lithium ions compared to the 0.5 to 0.7 for the material currently used to separate anode and cathode. Li-S batteries can be used for electric vehicles, but the technology could be adapted to portable electronic devices as well.
  2. Supercapacitors offer higher energy storage than conventional capacitors for applications such as burst or pulse load applications (e.g., LED flash), power amplifiers, specific audio circuits, and devices drawing very little current over a long time as in IoT. To overcome the supercapacitors’ 2.5-2.7 V range (opposed to Li-ion batteries’ 3.5-3.7 V range), supercapacitors can be connected in series, but complexity increases. Supercapacitors have also a low energy density: 10 Wh/kg against the 200 Wh/kg of Li-ion batteries.
  3. Fuel cells have been used to convert the chemical energy of fuel into electricity and are considered a good option for electric vehicles. For the time being, their size does not allow their use for electronic device batteries.
  4. NanoBolt lithium tungsten batteries can use a web-like nano structure obtained by adding tungsten and carbon multi-layered nanotubes that bond to the copper anode substrate thus enabling more ions to attach during charge/discharge cycles.
  5. Zinc-manganese oxide batteries can exploit the chemical conversion reaction in a zinc-manganese oxide batter and increase energy density in conventional batteries without increasing cost.
  6. Gold nanowire gel electrolyte batteries can use electrolyte gels – not as combustible as liquids – covering coating gold nanowires with manganese dioxide. The resulting electrodes can stand 200,000 charge/discharge cycles, instead of the 6,000 cycles of a conventional battery.

2      Consumption

The Metaverse is likely to cause a change of habits that will results in significant social savings e.g., by reducing the need for physical displacement and the use of real estate for commercial buildings. The importance of these savings can be assessed by the following data in the USA [44]:

  • Commercial buildings emit 826 MT/year of CO2 and consume 35% of the electrical energy.
  • To commute for 50 km by gas-powered car to those buildings a human emits 3.2 T of CO2/year.

The percentage of end use of energy by sector in AU, CA, CH, a selection of  EU countries, JP, KR, NZ, UK, US are:

Residential heating 11%
Other residential 9%
Industry 31%
Passenger cars 21%
Other vehicles 14%
Services 14%

The positive side of Metaverses is their potential to save a share of the currently consumed energy (and their future increases). The negative side is the energy cost of running big data centres containing Environments and Objects, transmitted over the Internet, and the Devices rendering complex Environments. An estimate of both factors is a necessary step before engaging in mass deployment of the Metaverse.

Table 4 gives some data related to global energy consumption (all values approximate) of some technologies expected to underpin the development of the Metaverse. The measure of Energy consistently used is kWh expressed with data ranging from 103 to 1018, according to the following conventional abbreviations:

kilo (k) =103 Mega (M) =106 Giga (G) =109 Tera (T) =1012 Peta (P) =1015 Exa (E) =1018 Zetta (E) =1018

Table 4 – Global energy consumption and data transmission (estimates)

Use type Target Data Unit Year Ref.
Energy supply All 165 PWh 2020 [46]
Energy supply Electricity 27 PWh 2020 [46]
Energy consumption All 111 PWh 2020 [46]
Energy consumption Electricity 23 PWh 2020 [46]
Energy consumption ICT 0.6 PWh 2017 [48]
Energy consumption Data centres 0.3 PWh 2021 [49]
Energy consumption Internet 0.2 PWh 2021 [49]
Data transmission Internet 1.2 ZByte 2016 [47]
Data transmission Internet 3.3 ZByte 2021 [47]

An initial analysis of some of the technologies to be taken into consideration because of their relevance to energy consumption is done in the following:

  1. Datacentres.
    1. A 2015 study before the advanced cloud gaming platforms like Google’s Stadia and Microsoft’s Xbox Cloud Gaming stated that datacentres were responsible for ~2% of global greenhouse gas emissions [43].
    2. A Meta data centre planned to be deployed in the Netherlands (and currently on pause) to host a portion of the metaverse in Europe has an expected energy consumption of 1.4 TWh/year.
  2. AI Model training.
    1. According to a 2019 study by the University of Massachusetts at Amherst, one AI language processing model’s estimated carbon footprint and electricity cost is >284 T of CO2 [43].
    2. Generative AI systems used to train AI models to generate a set of images given a text prompt, e.g., DALL-E[1], require a large amount of computing power. An idea of the power consumed by DALL-E can be derived from the fact that the system is a scaled-down version of the text-writing AI system GPT-3[2] trained on pairs of text and images drawn from the internet. To train GPT-3, 1.3 GWh were used and the training produced 0.5 MT of CO2.
    3. Nvidia reports[3] that its StyleGAN3 consumes 225MW to generate portraits of people that don’t exist.
  3. Blockchains.
    1. In a Bitcoin network that processes ~7 transactions/s[4], a single Bitcoin transaction consumes 9 MWh[5].
    2. According to the Cambridge Bitcoin Electricity Consumption Index[6], Bitcoin accounts for ~0.40% of the world’s total electricity consumption, and ~0.34% of the world’s total electricity production.
    3. Digiconomist[7] estimates that Bitcoin’s annual energy consumption has risen from 9.6 TWh in February 2017 to 73.2 TWh in January 2020.
    4. A 2021 study found that Bitcoin mining consumes around 91 TWh of electricity annually, i.e., 0.5% of all energy consumption worldwide and more than seven times the electricity used by all of Google’s global operations.
    5. A single Bitcoin transaction consumes more energy than 100 000 Visa transactions.
    6. According to ARK Invest[8], traditional banking emits ~1,4 MT of CO2/y and gold mining emits 144 MT while Bitcoin emits 61 MT, less than 5% and 45% of traditional banking and gold mining, respectively.
  4. Gaming.
    1. A 2020 Greening The Beast study[9] estimated that high-end gamers with state-of-the-art VR hardware, will spend as much as $2,200 over the course of five years on electricity and pump as much as 1T of CO2/y.
    2. University of Bristol (UK) found that, if just 30% of gamers using 720p or 1080p devices were to transition to cloud gaming by 2030, it would cause a 29.9% increase in CO2 If 90% of gamers moved to the cloud, it would increase gaming’s overall CO2 emissions by 112%.

The IT industry, however, is quite accustomed to facing challenges:

  1. In 2017, the IEA expected global data centre energy demand to grow by only ~3% to 200 TWh in 2020 because the growth for data centre services is offset by improved servers, storage devices, network switches and data centre infrastructure, including by the use of hyperscale data centres [39].
  2. A 2020 analysis[10] found that energy consumption for all datacentres rose less than 10% from 2010 to 2018 in spite of an increase of 2,600% in server, storage, and network workloads hosted by the cloud datacentres.
  3. Others estimate that this offset may not continue indefinitely because the most obvious energy savings are introduced first.
  4. Perspectives of cloud service providers
    1. Google has committed to operating on 24/7 carbon-free energy in all its datacentres by 2030.
    2. Microsoft intends to be “carbon negative” by 2030, which includes a plan to stop using diesel fuel in its datacentre generators by 2030.
    3. Amazon Web Services aims to power its operations with 100% renewable energy by 2025.
  5. There are less energy-consuming implementations of blockchains compared to Bitcoin. As an example, Ethereum after switching to a proof-of-stake consensus mechanism (i.e., using ETH instead of energy to secure the network) only uses ~0.0026 TWh/y across the entire global network. From these values it is possible to estimate[11] at 2.6 MWh for the network’s annual electricity consumption (September 2022), which corresponds to 870 T of CO2/y emissions.

[1] https://venturebeat.com/2021/01/05/openai-debuts-dall-e-for-generating-images-from-text/

[2] https://venturebeat.com/2021/06/01/microsoft-gpt-3-and-the-future-of-openai/

[3] https://www.reddit.com/r/MachineLearning/comments/q6ark8/r_stylegan3_aliasfree_generative_adversarial/

[4] https://www.investopedia.com/terms/b/bitcoin-mining.asp

[5] https://digiconomist.net/bitcoin-energy-consumption

[6] https://cbeci.org/cbeci/comparisons

[7] https://digiconomist.net/bitcoin-energy-consumption

[8] https://ark-invest.com/articles/analyst-research/bitcoin-myths/

[9] https://www.mic.com/impact/gamings-environmental-impact-is-bigger-than-you-think-21753800

[10] https://www.nytimes.com/2021/06/24/technology/computer-energy-use-study.html

[11] https://ethereum.org/en/energy-consumption/

<– Network    Go to ToC    Governance–>