Power, Pollution and the Internet
nytimes.com
nytimes.com
From an economic argument, data center operators have a profit motive. A profitable enterprise consumes less "resource" than it consumes (as long as they are not subsidized by the government). So it's obvious that they are doing something valuable for the economy regardless of the residual "wasted" resources.
Compare to other human activities, where "waste" can be identified. If you live in a home that has a bedroom that you only use 8 hours a day, do we decry the 66% of waste that the home represents?
How about comparing the total energy costs of buying a physical book in a bookstore to downloading it to a Kindle? I have no doubt the new mode represents a savings of several orders of magnitude less energy consumed. But Mr. Glanz wants us to be outraged that the power distribution network wastes 30% of the minuscule amount of energy required to send it to me over the Internet.
Had he not wanted to write a sensationalized piece, he could have been more helpful in explaining data centers on a human level. How many watts of electricity are being spent in data centers on my behalf? Rather than sensationalize a "300 million watts" number for Google data centers, he could have said Google expends 3 watts per US household...but that number is not going to scare anyone.
> The pollution from data centers has increasingly been cited by the authorities for violating clean air regulations, documents show. In Silicon Valley, many data centers appear on the state government’s Toxic Air Contaminant Inventory, a roster of the area’s top stationary diesel polluters.
There is some interesting information, but reads as if it wants to be some kind of scandalous exposé with all the off the record quotes.
I like the proposed solution though:
> "Some industry experts believe a solution lies in the cloud: centralizing computing among large and well-operated data centers."
This of course doesn't always meet needs. Sometimes you need bare metal and your own network for security or latency reasons like maybe in the finance industry where milliseconds matter.
Some companies, academic organizations and research groups have shown that vastly more efficient practices are possible, although it is difficult to compare different types of tasks.
The National Energy Research Scientific Computing Center, which consists of clusters of servers and mainframe computers at the Lawrence Berkeley National Laboratory in California, ran at 96.4 percent utilization in July, said Jeff Broughton, the director of operations. The efficiency is achieved by queuing up large jobs and scheduling them so that the machines are running nearly full-out, 24 hours a day.
Brought to you by the Times' best and brightest.