Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 



Forgot your password?
typodupeerror
×
Power IT

Price of Power in a Data Center 384

mstansberry writes "Much like the rest of the country, IT is facing an energy crisis. The utilities are bracing companies for price spikes this winter and according to experts and IT pros, those prices aren't going to come down any time soon. This is thefirst article in a four-part series investigating the impact of energy issues on IT."
This discussion has been archived. No new comments can be posted.

Price of Power in a Data Center

Comments Filter:
  • by Anonymous Coward on Thursday October 27, 2005 @05:22PM (#13892496)
    it was a extra $200 a month. ouchy.
  • by gtrubetskoy ( 734033 ) * on Thursday October 27, 2005 @05:24PM (#13892508)

    We have a page [openhosting.com] on our site with some calculations on how much energy is being saved because we're using Linux VServer and why dedicated servers are not environmentally-friendly (at least not with the current technology - this may change). The numbers are probably off a bit, but they give you some idea.

    Also the street price for a 20A circuit in a datacenter is $200-$300, while the cost of a megabit is $100 or less. So a rack of servers that requires two power circuits and pushes 3Mbps (not an unusual scenario) costs twice as much in power than in bandwidth.

    And here's another article [eweek.com] on this issue. And another [geek.com].

  • by nixer ( 692046 ) on Thursday October 27, 2005 @05:50PM (#13892709)
    Check the power numbers on Intel CPUs - they exceed 100w for all of the enterprise stuff. In a dense rack using blade servers the number of processors can exceed 100 (and there are configurations of around 200). Do the math and you see than this is 10KW. If you look at the specs for a typical set up, the input for this kind of kit is around 20 to 24KW - so around 40-50% of the heat is the CPUs.
  • by Sir_Eptishous ( 873977 ) on Thursday October 27, 2005 @06:13PM (#13892885)
    As a US citizen, I agree completely, but you must understand the cultural forces at work here.

    Since the mid-80's, when energy conservation/innovation became passe, and something to be ridiculed like an 8-track tape or bell bottom jeans, the American marketing and advertising paradigm has increasingly encouraged waste and consumption. Since the 90's, when the explosion of consumer electronics really took off, American consumers have been on a binge of gadgetry, all of which require electricity. Yes, I know I'm a hypocrite, because I too love the gadgets and have a home network, etc;

    If anyone here remembers not only the energy crisis of the 70's, but also the summer of 2001 in California, with the rolling black/brown-outs, then you will understand how critical this is.
  • SpeedStep (Score:3, Informative)

    by tepples ( 727027 ) <tepplesNO@SPAMgmail.com> on Thursday October 27, 2005 @06:31PM (#13893015) Homepage Journal

    The biggest help would be a power-saving feature for the CPUs that when idle they go into a sort of sleep mode and turn off some parts to save power, but I don't recall ever seeing this option on anything but Disk Drives.

    It's already here. Once all processes on a system are blocking (waiting for something else to happen), the kernel normally executes a HLT (halt) instruction that waits for the next interrupt. Intel processors have reduced power consumption when executing HLT for as long as I can remember. Intel Pentium M processors also have SpeedStep technology [bay-wolf.com]. When not overridden by software, SpeedStep cuts the multiplier and the voltage when available power is reduced.

  • Re:Unctuous (Score:4, Informative)

    by twiddlingbits ( 707452 ) on Thursday October 27, 2005 @06:41PM (#13893066)
    You forget that the US companies such as Exxon are partners with these National Oil Companies in many of these reserves. In fact the Saudi National Oil Company sold "concessions" to produce and market the oil to American firms as recently as 1998 for certain production.

    Saudi Arabs, Americans and Oil
    By Robert L. Norberg

    Human Resources

    In 1949, when Harry Snyder was hired to head up the training of Saudi Arabs for Aramco, James Terry Duce, a company executive in New York, told him what was expected:
    Your task at Aramco is to train Saudis as quickly and as soundly as possible to operate the Saudi oil industry. Inevitably, the Saudi Arab Government will eventually nationalize the industry. When that occurs, we want the young Saudis to have attained the proficiency that will enable them to operate the oil industry efficiently and with goodwill toward Aramco. Thus they will be serving their country's best interests and will be protecting the interests of our parent companies.1

    This vision of the training mission and its ultimate result might have appeared reasonably attainable if recruits were available from local schools, knew a bit of English, and had some exposure to industrial practices. But those conditions did not exist when the concession agreement was signed in 1933, nor in 1949 as the postwar development of Saudi Arabia's petroleum resources gathered momentum. Tom Barger, a geologist who arrived in Arabia in 1937 and rose to board chairman before retiring in 1969, recalled many years later:
    [One] aspect that impressed me was the enormous, inordinate poverty of the inhabitants. As I found out later, nearly everybody was hungry most of the time. . . . There's no education, obviously. The few people who could read and write largely had taught themselves. And there were some very learned men, as a matter of fact, among this population, although most of it was illiterate. They had practically no mechanical skills. We had new employees who couldn't get out of a room because they didn't know how to use a doorknob."2

    B. C. Nelson, who served Aramco in employee relations for many years, recalled in 1965 what it had been like for Saudis recruited to Aramco in the early years of the enterprise:
    Word spread to the desert and townspeople that in exchange for some physical effort the blue-eyed foreigners would give a man a handful of silver! And so they flocked to Aramco's budding oil centers . . . Imagine the effect on a recruit to be plunged into the mechanical age -- none of which fit in with his prior orientation or culture -- with little or nothing in his experience to help him adjust. The most amazing thing about these times in terms of one small facet of an Industrial Relations problem -- absenteeism-was not that, when they were handed their bag of money, they returned to their tribe with their glad tidings, but rather that they ever came back to work. Industrial discipline was practically unknown, so the amazing thing was that there was only a 75 percent turnover in the first few years.3

    On-the-job training began on an informal basis in the 1930s and was soon complemented by rudimentary industrial training in classrooms. But without English, Arabic literacy, and basic arithmetic, there was a limit to the progress Saudis could make in job performance and advancement. In 1944, with operations revived after a wartime suspension, the Jabal (meaning "mountain" or "hill") School was opened in Dhahran.
    Surely in 1944 no one expected history to remember the humble Jabal School. Yet the little company school endures as a symbol for development -- not for the development of an oil company, but for the development of a generation of very special young men. Many Saudis were introduced to the mystery of letters and numbers at the Jabal School. Among them were future scholars, successful businessmen and powerful executives.4

    The Jabal School was the beginning of an ever-evolving, structured program of job-related training and general education that replicated under corpor
  • by Trogre ( 513942 ) on Thursday October 27, 2005 @07:30PM (#13893360) Homepage
    I believe it's not so much a high temperature that affects electronic equipment, but fluctuations in temperature.

    If your PCB traces are expanding and contracting between say 20C at night and 40C during the day you're going to get fatigue. It's also not so good for mechanisms inside hard disk drives.

    So the HVAC guy at the local television studio tells me anyway.

  • by colinnwn ( 677715 ) on Thursday October 27, 2005 @07:43PM (#13893438)
    There is really no short-term method governments can control "price gouging" outside of price caps (which is fixing an upper limit on price). Long-term the real issue is caused by the target industry's regulations which create artificial entry barriers.

    People use the term "price gouging" anytime they percieve the price of a good is too high. This is a fallicy. The definition of price gouging is http://dictionary.reference.com/search?q=price+gou ging [reference.com] pricing above the market when no alternative retailer is available. There are plenty of alternative wholesalers and retailers in the global and domestic oil and gas market. The only real reason for lack of alternative retailers would be when government regulation impedes market entry.

    I would argue what people percieve as "price gouging" is actually beneficial to a majority of consumers. Lets take the gas situation in Louisana immediately after Katrina as an example. The demand for gas increased substantially due to evacuation and rebuilding efforts. In a free market, this would quickly drive the price of gas well above its production costs. Louisana has "price gouging" laws, so retailers were not able to price gas at market levels. The first few consumers bought all the "cheaper" gas they could carry away and everyone else got nothing. None of the retailers were interested to hang around in the bad conditions and try to acquire and sell more gas. They knew they wouldn't earn substantial profits doing so. They got out of town, to come back and sell gas when conditions were better and it was more easily available.

    If those retailers had been able to price gas at market levels, the first few consumers would have purchased only the gas they needed and not all they could get away with. The next group of consumers would have been able to acquire some much needed gas also at more expensive prices, instead of getting none. When the gas in storage was gone, those retailers would have looked at their pile of cash and said, "My goodness I like this. I am going to stick around in the miserable conditions and do whatever it takes to get more gas in here to sell." Now the 3rd group of consumers that got no gas with "price gouging" laws would be able to purchase some newly delivered and even more expensive gas. I think most consumers that got no gas would have been willing to pay a lot more for a little bit of gas.

    "These prices are high because of risk, not insufficient supply."
    Risk is priced into the supply curve that shows how much producers are willing to supply at each price point. If risk increases, it pulls the supply curve in. Producers are willing to supply less at any price, moving the quantity demanded lower and price higher back into market equilibrium. It is inherent to the market economy and most feel it is much better than a government managed economy. Think of how much dispute and consternation is put into political process in this country; now imagine if the same thing happened with every economic production and sales decision.

    "China, our enemy."
    You might want to rethink this. We might not agree with China's political decisions right now, but the only reason the US is not taking a harder stance with China is they are our best friend and savior economically. They produce commodity goods for us much more efficiently than we can, allowing us to buy more than we otherwise could and keep our standard of living higher. They are the largest holder of our currency, keeping its value stable enough to remain the world standard currency. They also finance our obscene deficits both public and private, allowing us to keep our economy and the world economy out of recession.
  • Re:Unctuous (Score:3, Informative)

    by Jherek Carnelian ( 831679 ) on Thursday October 27, 2005 @07:43PM (#13893439)
    How do you know it's fuelled by gouging and speculation? How can you say the price is "wrong"? Define "gouging".

    Whatever the cause it is evident that something is wrong with the market. As others have pointed out, the oil companies are reporting overwhelmingly huge profits. In an efficient market, cost of production and selling price converge. Since they are currently hugely divergent, somebody is jiggering the system. Since the USA, and anything to do with oil the world around, is pretty much a corptocracy, chances are the blame lays with the corps.
  • by spatenbrau ( 926486 ) on Thursday October 27, 2005 @09:21PM (#13893965)

    Under linux the laptop was drawing 50-100watts (which is very high for a laptop), under windows it was drawing from 30-50 watts.

    There are several aspects to power managment that many of the OSS kernels fail to take advantage of. 1) they don't turn down they CPU voltage and clock rate, even when the CPU is idle. 2) they don't turn the power off to unused chips and peripherals. 3) sometimes they don't even turn off some devices when the laptop is turned off. Anyone with a Compac Evo laptop has probably seen this. Shutdown your computer while running linux on mains, unplug it so you can take it home and look at the battery charge a few days later and it is down to half charge. Some of the internal devices (like ethernet transcivers) are never even turned off. While wake-on-lan etc. might make sense on desktops, one has to wonder what the folks were thinking when they decided to keep a laptop's ethernet powered even when the laptop might later be needlessly draining the battery.

    Now some of this isn't the the fault of the OSS authors. In many cases the chip companies simply never release the information needed to correctly manage the power. That's got to change. Other times, it is simply that nobody cares enough about power usage to get off their butt and write the code. Certainly the amd64 CPU voltage/frequency settings fall into that category for the various BSD's. How many amd64 computers are out there running OSS, idling at 140watts when they could easily be idling at 90watts, a savings of ~30%.

  • by Rei ( 128717 ) on Thursday October 27, 2005 @10:03PM (#13894123) Homepage
    Given that every bit of data that they've gathered so far fits into its predictions from the previous data within the range that they would expect it to hold up over given the resolution, for the past century, I'd say that forecasting global trends for another thirty years is well within their range.
  • by maxwell demon ( 590494 ) on Friday October 28, 2005 @05:41AM (#13895343) Journal
    Erasing a bit at 300 K (27C or 80F) theoretically costs a minimal energy of 2.87e-21 J. That is, an ideal non-reversive computer at 3 GHz erasing 1 GByte at each cycle would just need about 70 milliwatt. It's obvious that our current computers are very far from that limit.

    Also note that current power supplies are all but efficient. That is, a lot of the energy your computer draws from the grid doesn't even reach the CPU.

    Since our main losses are obviously not the inevitable cost of non-reversible computing, but other losses dominate, reversible computing will clearly not be the solution to the current computing energy problems.

He has not acquired a fortune; the fortune has acquired him. -- Bion

Working...