Energy consumption of electronics

From ScienceForSustainability
Revision as of 17:40, 11 February 2021 by Sisussman (talk | contribs) (→‎How it works)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search


Computers, phones, the internet, data centres and the applications that run on them all consume energy. Some, like electronic currencies such as Bitcoin, can consume a lot. Others, such as mobile phone chargers left connected to the mains when the phone is unplugged, consume practically nothing. This article discusses a couple of cases.

Bitcoin

Bitcoin and other electronic currencies use virtual coins which are obtained by solving numerical puzzles using computers. In order to keep the rate of creation of new coins constant as more participants attempt to create them, the difficulty of the puzzles, and the amount of computing effort required to solve them, increases. The amount of electrical power used by participating computers is a significant proportion of global energy consumption - in the tens of TWh per year - equal to that of many countries.

Bitcoin Energy Consumption Index Digiconomist

Ever since its inception Bitcoin’s trust-minimizing consensus has been enabled by its proof-of-work algorithm. The machines performing the “work” are consuming huge amounts of energy while doing so. The Bitcoin Energy Consumption Index was created to provide insight into this amount, and raise awareness on the unsustainability of the proof-of-work algorithm.

Rebuttal to Digiconomist comparison between Bitcoin and Visa:

The Bitcoin vs Visa Electricity Consumption Fallacy Carlos Domingo; Hackernoon; 30 Nov 2017

Bitcoin is not simply a piece of a payment network like VISA but a full currency system, VISA itself requires the banking system for its payment system to work so you need to actually include some of those costs there to make a meaningful comparison.

Ethereum Energy Consumption Index Digiconomist

Ethereum has plans to change its proof-of-work algorithm to an energy efficient proof-of-stake algorithm called Casper. This change would minimize energy consumption and will be implemented gradually according to the latest roadmap. For now, Ethereum is still running on proof-of-work completely.

Bitcoin could cost us our clean-energy future Eric Holthaus; Grist; 5 Dec 2017

Bitcoin’s energy usage is huge – we can't afford to ignore it Alex Hern; The Guardian; 17 Jan 2018

Bitcoin’s electricity usage is enormous. In November, the power consumed by the entire bitcoin network was estimated to be higher than that of the Republic of Ireland. Since then, its demands have only grown. It’s now on pace to use just over 42TWh of electricity in a year, placing it ahead of New Zealand and Hungary and just behind Peru, according to estimates from Digiconomist. That’s commensurate with CO2 emissions of 20 megatonnes – or roughly 1m transatlantic flights.

That fact should be a grave notion to anyone who hopes for the cryptocurrency to grow further in stature and enter widespread usage. But even more alarming is that things could get much, much worse, helping to increase climate change in the process.

The Carbon Footprint of Bitcoin by Christian Stoll, Lena Klaaßen, and Ulrich Gallersdörfer in Joule on 12 June 2019 [article] (full paper paywalled)

Highlights

  • Bitcoin's annual electricity consumption adds up to 45.8 TWh
  • The corresponding annual carbon emissions range from 22.0 to 22.9 MtCO
    2
  • This level sits between the levels produced by the nations of Jordan and Sri Lanka

Summary

Participation in the Bitcoin blockchain validation process requires specialized hardware and vast amounts of electricity, which translates into a significant carbon footprint. Here, we demonstrate a methodology for estimating the power consumption associated with Bitcoin’s blockchain based on IPO filings of major hardware manufacturers, insights on mining facility operations, and mining pool compositions. We then translate our power consumption estimate into carbon emissions, using the localization of IP addresses. We determine the annual electricity consumption of Bitcoin, as of November 2018, to be 45.8 TWh and estimate that annual carbon emissions range from 22.0 to 22.9 MtCO 2. This means that the emissions produced by Bitcoin sit between the levels produced by the nations of Jordan and Sri Lanka, which is comparable to the level of Kansas City. With this article, we aim to gauge the external costs of Bitcoin and inform the broader debate on the costs and benefits of cryptocurrencies.

How it works

In a Facebook post science communicator Zi Teng Wang wrote:

In 2008, a paper was published by a person or group of people under the pseudonym Satoshi Nakamoto. This now-legendary paper proposed an elegant solution to a long-unsolved problem in computer science called the Double-Spend problem.

Imagine a token which symbolizes value. It could be anything, but common tokens we use today include pieces of paper with pictures of old white men, numbers in computer databases which represent those pieces of paper, bits of specific metals pressed into shapes, and rare but pretty rocks.

Imagine that someone wants to give you some tokens, in exchange for something of value that you provide them. How do you protect yourself from being cheated, how do you guarantee that those tokens will remain yours?

With physical tokens like cash, metal, or gems, it's easy: it can only exist in one place at a time, so as long as you're holding onto it, it can't be anywhere else. The major downside is that they can only be transferred in person, and transferring large amounts of it at a time can quickly get hairy.

But electronically? Anyone in the arts and media world can tell you that things which exist only electronically can be trivially duplicated and reproduced. So if someone is sending you a digital token online, how can you trust that you are now the only unique holder of it, and it isn't simultaneously being copied and sent to other places online?

The traditional solution is to defer the job of verification to a central authority, usually a bank. The bank keeps a ledger, a master database of who owns what. That person would notify their bank that they'd like to transfer ownership of some tokens to you, the bank would check their ledger, verify that they have it, then create a new entry into the ledger recording the transfer. If that person were to try and send the same tokens to someone else later, the bank would say no, you can't do that.

This works well in protecting you from needing to trust the other person, but it introduces a third party into the transaction, the bank, which you need to trust instead. A bank which needs to monitor and track what you have, always stands between you and the person you're sending tokens to, and who has the power to deny or even reverse transactions, and freeze or seize tokens and edit the ledger as it chooses. We put a lot of trust in banks not to do that, and they almost never do. But is it possible to create a system that is secure without requiring trust at all?

How can you create a ledger for a digital token which cannot be duplicated or counterfeited, which can only exist in one digital pocket at a time, which can be transferred from user to user freely but cannot be double-spent, where not even a single entity needs to be trusted with unilateral power over the books?

What Nakamoto proposed is something called a Proof-Of-Work Blockchain. A blockchain is a special type of ledger, maintained by a decentralized, trustless swarm of competing agents, which will nevertheless converge upon one single un-alterable consensus ledger of transactions because of the rules about how new entries to the ledger must be written.

In 2008, Nakamoto published their paper, and in early 2009 they released the first public blockchain onto the internet, and the tokens of this blockchain are called Bitcoins.

In order to understand how the blockchain works, there are two key concepts you need to understand: 1. Hashing algorithms, and 2. Public-Private Keypairs. Let's go over them now.

A hashing algorithm is a complicated bundle of math which functions like an extremely precise woodchipper. If you feed something in, it takes your input and shreds it into a string of digital static called a hash. If you feed exactly the same thing in, you'll always get exactly the same string out. If you change the input even a tiny bit, the hash that comes out will be completely different. And although it's easy to shove an input in and get a hash, it's impossible to take a hash and use it to reconstruct the original input.

A Public-Private keypair is a secure way to digitally prove your identity and ownership. The Public key can be disseminated widely, and the Private key is kept secret by the owner. The Private key can be used to prove that you're the owner of the Public key, and the Public key can be used to encrypt messages that are only decryptable by the holder of the Private key.

Like any bank's internal ledger, the Blockchain is a massive database, a huge list of public wallet addresses (corresponding to public keys) along with data on how much bitcoin each one contains. Alongside that, it also contains the complete transaction history of every single transfer of bitcoin from one address to another since the blockchain was first created in 2009. This file is understandably massive- over 300 gigabytes as of Jan 2021. True to its name, each new batch of transactions added to this database is called a block, and each block is built off the previous one like links in a chain.

However, unlike the ledger of a bank, which the bank can edit and change with ease, adding a new batch of transactions to the Blockchain ledger is an intentionally hard process, and the deliberately wasteful difficulty of doing so is the key to its security.

Suppose you wanted to receive bitcoins from someone. You'd randomly generate a public-private keypair, and you'd give them the public key you just generated. They would have the private key for a public wallet address which exists on the blockchain and is registered as having some amount of bitcoin, and they would use their private key to transmit a message, "hey, wallet XYZ transfers 0.22524511 bitcoin to wallet ABC, here's proof that I'm the owner of XYZ. By the way, I'm including a tip of 0.00005000 bitcoin to the Miner who processes this".

This message goes to what is essentially a public noticeboard (called the mempool), alongside other messages from other people in the world who also want to record a bitcoin transaction into the ledger.

Then, the Bitcoin Miners come in. Remember the decentralized swarm of agents we mentioned earlier? Bitcoin Miners are the people (technically, the computer systems they set up) who monitor the mempool and are in constant, furious competition to luck out and become the author of the next block, and it is through their efforts that everyone's transactions get written into the blockchain. The reward for doing so is that, in addition to the tips collected from each transaction, each block written allows the author to declare the creation of a new public wallet address which has coins from nowhere, to which the miner holds the private key.

A block might look like,

-------------------------------
The previous block was block #149
This is block #150
XYZ transfers 0.22524511 coins to ABC and 0.00005000 coins to MYN
SRM transfers 15.4250000 coins to KJQ and 0.00004500 coins to MYN
JOE transfers 0.00752000 coins to LEA and 0.00002200 coins to MYN
The new address for this block is MY2, which now contains 50.00000000 coins
End of Block #150
Zombie Pirate Party Monkey
-------------------------------

You might have 2 questions in mind right now:

  1. . What's stopping a miner from just writing blocks as fast as they want and minting coins for themselves as fast as they'd like?
  2. . What's the deal with the last part of that block?

The answer to both questions are the same, and they're related to Hashing. If you take the entire block and shove it through a hashing algorithm, you'll discover that the hash which comes out looks extremely unusual.

This is the hash for the real block #150 on the actual blockchain (and its successors)-

#150- 000000009ca75733b4cf527fe193b919201a2ed38c9e147a5665fdfade551f4d
#151- 00000000f04ec51395de63f4c3c76766d012ce73eaabe7eceaf124eb7696f36a
#152- 00000000ab6990d52120e8db495dffb69c995b9091ad2424a5bfc934e04462c6

Notice a pattern? The critical rule of Bitcoin mining, which stops miners from just writing as many blocks as they'd like, as fast as they'd like, is this: the hash of each block must start with a certain amount of zeroes.

In order to write this block, the miner first writes a draft: "The previous block was #149, this is block #150, people sending coins to each other etc etc, tips go to my wallet which is here, and here's my new wallet with 50 coins I want to add to the blockchain", and then they hash it. It probably won't start with the needed amount of zeroes. So they add a random bit to the end and hash it again. Still no. Different bit, no. The miner will keep throwing different random bits into their draft block, billions of times per second until, by pure luck and brute force, they hit upon a random string which makes the block pass.

"Eureka! Here's Block #150!" they announce, broadcasting the new block to other miners. The other miners, busy with their own billions of guesses per second, spare a single hash cycle checking out this #150, confirm that the hash checks out, drop their own personal drafts of #150, adds this #150 to their own personal copies of the blockchain and start working on block #151, hoping to write the next block themselves this time.

As the miners spread word of the new block #150, you're satisfied to see that person's transaction now recorded onto this latest link of the blockchain. Think about what would need to happen if, thirty minutes later, the person who sent you those coins wanted to take it back, erase that record, and double-spend it elsewhere. What would they have to do?

They would have to write an alternate block #150, one that didn't include their transaction to you. In order to do that, they would have to make trillions of guesses themselves before they found a random string that would make it hash properly. Then, they'd have to write an alt-#151, alt-#152, and so on. Meanwhile, the majority of the miners have already published block #153 and are now working on block #154, and nobody would accept their alt-#150 because everyone is already racing to be the one who hits the jackpot on writing block #154. The only way for a nefarious agent to reverse and double-spend a transaction which has already been written into the consensus chain would be to create an alternate chain longer than the dominant one, and doing that would require commanding more computer power than the rest of the network combined. Yeah, good luck on that. By the time a few more blocks get written on top of the block with your transaction of interest in it, that transaction is, for all practical purposes, permanently set and irreversible.

And that's how you solve the double-spend problem and create a digital token that can't be duplicated or counterfeited, can only exist in one place at a time, whose supply is finite, which can be sent freely from any address to any other, and is secured not by trusting a third-party authority, but by cryptography and math.

It's hard to explain and hard to compare to other things because honestly, until that day in 2009 when Satoshi Nakamoto first launched the bitcoin blockchain onto the open internet, nothing exactly like this has ever existed before.

At its core, this is what Bitcoin is, and what it continues to be today.

Built around that core are the far more messy human and economic questions of, what is it worth? How much are people willing to buy and sell it for today, and what price will they be willing to pay for it tomorrow? These questions and the rancid sewers of discussion around them are beyond the scope of this technical dissection, but I will point out one observation that's always struck me as particularly interesting.

Bitcoin came into existence early in 2009, and at first it was largely just a curiosity and a plaything for cryptography nerds, computer scientists, and radical economic libertarians. People connected their computers to the mining network just for fun and curiosity's sake, made wallets and transferred buckets of bitcoins around and sent them to friends and gave them away to random strangers just because they could. It was a pretty neat toy, but the big question at the time was, "Okay, yeah it's cool, but does this stuff actually have any value at all?"

Eventually, on May 22, 2010, mostly on a lark, a Florida man paid another person 10,000 bitcoins for them to remotely order him two Large pizzas from Dominos. By this act, they became the first two people in history to believe that the answer to that big question is: Yes, it does.

So far, people have not stopped believing that yet.

Tesla

In February 2021 Tesla invested $1.5bn in Bitcoin. In a post on LinkedIn Laszlo Varro, Chief Economist at International Energy Agency, calculated:

Tesla invested 1.5 billion to bitcoin. I could not find a detailed disclosure, but lets suppose the 1.5 billion was 37500 BTC at a 40000 dollars average price. At a conservative, 0.15 GWh/BTC mining electricity use this is 5625 GWh for mining Tesla’s bitcoins, roughly equivalent to the annual power supply of the 25 million residents of Cameroon. Tesla’s renewable business last year deployed 0.2 GW solar. At a normal solar load factor this is 240 GWh/year. It will take 23 years for the Tesla solar panels to compensate for the energy footprint of their bitcoin.

Netflix

Articles circulating on the internet in early 2020 claim that watching Netflix has a high carbon footprint.

An article, Factcheck: What is the carbon footprint of streaming video on Netflix? by George Kamiya in Carbon Brief on 25 Feb 2020 [article] examines these claims:

The use of streaming video is growing exponentially around the world. These services are associated with energy use and carbon emissions from devices, network infrastructure and data centres.

Yet, contrary to a slew of recent misleading media coverage, the climate impacts of streaming video remain relatively modest, particularly compared to other activities and sectors.

Drawing on analysis at the International Energy Agency (IEA) and other credible sources, we expose the flawed assumptions in one widely reported estimate of the emissions from watching 30 minutes of Netflix. These exaggerate the actual climate impact by 30- to 60-times.

The relatively low climate impact of streaming video today is thanks to rapid improvements in the energy efficiency of data centres, networks and devices. But slowing efficiency gains, rebound effects and new demands from emerging technologies, including artificial intelligence (AI) and blockchain, raise increasing concerns about the overall environmental impacts of the sector over the coming decades.

Misleading media

A number of recent media articles, including in the New York Post, CBC, Yahoo, DW, Gizmodo, Phys.org and BigThink, have repeated a claim that “the emissions generated by watching 30 minutes of Netflix [1.6 kg of CO2] is the same as driving almost 4 miles”.

The figures come from a July 2019 report by the Shift Project, a French thinktank, on the “unsustainable and growing impact” of online video. The report said streaming was responsible for more than 300m tonnes of CO2 (MtCO2) in 2018, equivalent to emissions from France.

The Shift Project’s report continues to influence media coverage, including articles published earlier this month by the Guardian and Thomson Reuters Foundation.

The Shift Project’s “3.2kgCO2 per hour” estimate is around eight times higher than a 2014 peer-reviewed study on the energy and emissions impacts of streaming video.

That 2014 study found streaming in the US in 2011 emitted 0.42kgCO2e per hour on a lifecycle basis, including “embodied” emissions from manufacture and disposal of infrastructure and devices. Emissions from operations – comparable in scope to the Shift Project analysis – accounted for only 0.36kgCO2e per hour.

Because the energy efficiency of data centres and networks is improving rapidly – doubling every couple of years – energy use and emissions today will be even lower.

Looking at electricity consumption alone, the Shift Project figures imply that one hour of Netflix consumes 6.1 kilowatt hours (kWh) of electricity. This is enough to drive a Tesla Model S more than 30km, power an LED lightbulb constantly for a month, or boil a kettle once a day for nearly three months.

With users collectively watching at least 165m hours per day, the Shift Project figures imply that Netflix streaming consumes around 370 terawatt hours (TWh) per year, which would be comfortably more than the annual electricity demand of the UK.

For comparison, these figures are 800-times larger than figures reported by Netflix (0.45TWh in 2019) and nearly double the estimated electricity use by all data centres globally (198TWh in 2018). It is clear that the Shift Project figures are too high – but by how much?

read more (full article includes links)