We know three things about transistors.
* They have already changed the world.
* Silicon technology progress is not over, and vast domains of untapped social and economic opportunity remain.
* Every single transistor consumes electricity.
And we know two other related facts:
* The energy cost of computation has collapsed, but the overall amount of energy consumed for computing and communicating has risen dramatically, and thus;
* The cost and availability of electricity to fuel the information-centric part of the world's economies are of ever-increasing importance – and that's where coal comes in.
The future demand for coal, for any primary fuel source, is, to state the obvious, fundamentally anchored in electricity demand. How much more will be needed, at what quality (specifically the quality of reliability), and at what price?
And the future demand for electricity is rooted in just two variables:
* Demographics — population, wealth, economic & social activities, and population concentration
* Technology — the invention of new things, and innovating existing things that consume electricity.
Information technology has had and continues to be an integral feature of all aspects of the above factors.
Information technology is now an enormous and fast-growing part of the economy.
In today's America, the $1.1 trillion share of our economy devoted to information and ideas — moving bits — has surpassed the share associated with transporting people and stuff. The gap will widen. This is the future as well for developing nations.
The atoms-moving industry is almost entirely about liquid fuels, where oil dominates — the bits-moving industry is all about electricity, where coal dominates.
Making and moving bits involves everything from digital movies and server farms, from iPhones and Intel to Netflix and Facebook, all with much more growth potential than all aspects of moving stuff and people, the classic transportation sector, which includes everything from making and using cars and airplanes, from F150s to Norfolk Southern and FedEx.
In the decade following 2000, the transportation sector of the U.S. GDP grew 15%, while the information sector grew 45%.
The rise of the Information Economy is driven primarily by the historic, and on-going, radical drop in computing costs.
Demand follows price.
Information processing technology has seen a decline in real costs that is unprecedented in the history of any product or service – and it continues. You would have had to spend $10,000 in 2000, $100 million in 1980, and $10 billion in 1960 to buy an iPad's computing capability.
Computing and communications technologies now enter a new phase that Morgan Stanley calls the Mobile Internet.
The Next Explosive Phase of the Information Economy Begins
The new era is enabled by four intertwined macro trends:
* The Cloud
- Centralized computing power expands exponentially in the Cloud, already comprised of tens of thousands of data centers any one of which has trillions of times the compute power of old mainframes or modern laptops.
* Wireless broadband
- Wireless networks, launched in 1980, are now one million times faster and bandwidth costs have dropped 100-fold.
* Smart phones
- Since 1980, computing speeds have risen 200 thousand fold; costs dropped one million fold enabling smart phones and tablets more powerful than 1980 mainframes..
* Smart things
- By 2003 more things than people on Earth were connected to the Internet. Within a decade IBM and Cisco forecast a trillion things will be connected. Any object, from lights and shoes, to doors and sphygmomanometers can have embedded logic and wireless connections.
Rising wealth is highly correlated with greater penetration and use of information & communications technologies (ICT). The United Nation's International Telecommunications Union ICT Development Index (IDI) is a composite of 11 indicators capturing the various aspects of the deployment and availability of ICT.
The future prosperity for emerging economies is now no longer measured in just light-bulbs-per-capita (though that is still vital, and a vastly under-served market for billions of people), but also in bits-per-capita.
And while light bulbs, refrigerators and water pumps require massive power plants, information technologies require massive data centers that in turn require power. Lights require the Grid. Transistors require the Cloud, and more Grid. And more coal — given the remarkable scale of demand created by more-transistors-than-grains-of-wheat.
Electric Demand Rises to Fuel the Cloud's Data Centers
Data-center energy consumption already rivals energy use associated with global aviation. And while computing and storage technology efficiencies will continue to improve, total data center electricity use is forecast to grow rapidly.
By 2020, the world's data centers alone will consume over 1.5 trillion kWh of electricity. That totals more than the countries of Japan and Germany combined. According to the International Energy Agency forecast, that amount of electricity will be supplied by almost 500 million more tons of coal per year.
The information economy consumes electricity directly and indirectly in numerous other ways not counted in the above graph: in the communications networks, in the manufacturing of transistors and information appliances, at the locations where consumers and businesses use that equipment, and in the IT-driven economic expansion of associated non-IT equipment in business and industries. All subjects for another day.
1 Note: more detailed unbundling would expand the IT share of the economy; other 'buried' IT-centric activities should be extracted that are otherwise counted separately in the Census; e.g., for IT aspects of government and similar services, for air traffic control (counted as "transportation"), or for the share of electrical equipment manufactured that is exclusively used in powering ICT equipment, etc.
The Big Sky Business Journal
P.O. Box 3262
Billings, MT 59103