August 3, 2022

More Bits, Fewer Watts

By
Jim Harris
Magazine Article

Energy Efficiency must be seen as the First Fuel Mobile Operators Must Prioritize Energy Efficiency

Digital services data traffic will grow by 13-fold between 2020 and 2030. Without energy efficiency improvements, the industry's energy consumption and carbon dioxide emissions will grow 2.3-fold, according to third party research. (name the 3 party) 

The Information and Communications Technology (ICT) industry will have to cut carbon emissions by at least 45% according to the United Nations Framework Convention on Climate Change (UNFCCC) Paris Agreement to fight the climate crisis. 

Energy efficiency must be the first fuel. 

Image

With more computing power, we need more efficiency: “More bits, fewer watts” was one of the key messages at Huawei’s Innovation Week held online and onsite in Shenzhen, China between July 18-21, 2022. 

Ryan Ding, President of Huawei's Carrier Business, stressed operators need to prioritize energy efficiency and collaborate on global standards for the industry. 

During the event, temperatures in Spain hit a record 43°C, and 40.3°C in Britain, (beating the British record by more than one full degree). At the same time, France was experiencing widespread wildfires. 

“We are facing an unprecedented challenge,” said Ding. 

“As more and more industries are going digital, the demand for data will rise sharply, which will result in a surge in energy consumption. In the meantime, the whole world is working to combat climate change, and the ICT industry must urgently achieve carbon peak and carbon neutrality.”

Just looking at this challenge it might seem hopeless, history however has shown that energy intensity industries can experience massive energy efficiency improvements. Take, for instance, lighting: 

Image

Radical decline in lighting costs from £40,000 ($US50K) in 1300 to basically free today with LED bulbs. In fact, today, lighting is so cheap that we take lighting for granted. 

Image: https://ourworldindata.org/grapher/the-price-for-lighting-per-million-lumen-hours-in-the-uk-in-british-pound

Most people have heard of Moore’s Law, coined by Gordon Moore in 1965. He said that the number of transistors on a CPU (central processing unit) of a computer would double every 24 months while staying at the same price point. The result of this when viewed over time is radical and revolutionary. In the early 1970s, an Intel microchip had 2,000 transistors on it. Today chips have 50,000,000,000 transistors! 

Image

Image source: https://ourworldindata.org/uploads/2019/05/Transistor-Count-over-time-to-2018.png

That, in turn, has had a radical impact on the cost of computing. A gigaflop is a measure of compute speed: it’s doing a billion transactions in a single second. In 1961, a gigaflop cost $165 Billion on mainframe computers. In 2022, it cost a single cent: 

The result is that computing power is basically free (by comparison to 1961) and it’s at the edge, meaning on my smartphone. (My smartphone has as much raw computing power as IBM’s Deep Blue project that beat Garry Kasparov in chess in 1997. That project cost $10 million while by comparison my smartphone is worth $1,000). 

Image

Today, our computational use has grown exponentially. Blockchain and crypto currency mining is measured in TH/s – Terra Hashes per second – or TRILLION calculations per second: 

Image

Image source: https://www.blockchain.com/charts/hash-rate (max time frame) 

Total global mining power hit 1 million TH/s in for the first time. Today it's 205 million TH/s. That means globally all the blockchain and crypto mining servers are performing 205 million trillion calculations every second! 

This exponential increase in total mining operations is the bad news. The good news is that energy efficiency has improved exponentially as well. The International Energy Agency (IEA) analysis is that mining operations are 3 MILLION times more efficient today than originally in 2010: 

Image

Image source: https://medium.datadriveninvestor.com/bitcoin-optimizes-energy-use-conversion-a25297cd765b 

Some people mistakenly feel that energy efficiency costs too much. 

Here are two personal examples: In 2004 I bought a Toyota Prius for $30,000. It used half the fuel of my prior vehicle. Over the 17 years that I owned it; this hybrid electric vehicle saved me more than $35,000 in fuel costs! 

There is a relationship between CAPEX (capital expenditure or the cost to buy something) and OPEX, the operating costs over the lifetime of the assets. Sometimes energy efficient technologies can cost a little more upfront, but the savings over the lifetime can be huge. 

Automakers in Detroit for decades have resisted energy efficiency standards for vehicles. If every car in the US had the same fuel efficiency as the Prius, America would not have had to import any foreign oil. 

In other words, the refusal to aggressively embrace energy efficiency in transportation placed a voluntary (and unnecessary) fuel tax of hundreds of billions of dollars a year on the American public. 

Lighting 

Since I began this article focusing on lighting, I am going to come back to it. 

Many years ago, I replaced most of the light bulbs in our home with LED lights. I bought the bulbs when the local electric utility had deep discount promotions on buying the bulbs. I got them for about $1 each. Here’s the economic case of replacing 40 incandescent bulbs with LEDs: 

A typical 100-watt incandescent bulb will cost 50 cents and last 750 hours. So over 10,000 hours of lighting here is the CAPEX and OPEX analysis: 

Each bulb will use $130 of electricity, and 13.3 bulbs will be required at a cost of $6. 67. So the total cost will be $136.67. 

By contrast the LED bulb last for 10,000 hours and will use $19.50 of electricity. 

The difference is $116.70 per bulb. Given that I changed 40 bulbs the total savings over the 10,000 hours will be $4,647. That’s an incredible financial return!  

But the analysis doesn’t take into account the cost of continually driving to the hardware store to buy new lightbulbs as the old ones burn out or the time (and risk) to get out a ladder to replace a bulb in a high and hard to reach place. 

Misaligned Incentives 

Typically, the construction cost for a new building is 30% of the lifetime cost of the building. By contrast the operating costs will represent 70% of the cost over the lifetime of a building. Spending 1% more on construction can save 50% of the operating costs or more. 

The problem? Someone in the purchasing department is given a bonus for getting the lowest price possible on construction, saddling the organization with high lifetime operating costs. 

Summary

ICT operators and mobile carriers need to look at the lifecycle costs of technology and invest in technologies that provide the best energy efficiency and the lowest life cycle costs – because the climate crisis demands radically improved energy efficiency from the industry. Thankfully it’s possible! 

Jim Harris is a futurist & professional speaker. On Twitter, @JimHarris was the most influential personal account for CES 2022. You can email him at jim@jimharris.com

Recent Posts

Tesla’s market capitalization (value) is greater than the combined value of General Motors, Ford, Fiat Chrysler, Toyota, Honda, Volkswagen, Nissan, Daimler (Mercedes Benz), Hyundai, Kia, BMW and Renault as of January 2022.

Continue reading