Slashdot Log In
DC Power Poised To Bring Savings To Datacenters
Posted by
timothy
on Wed Jan 14, 2009 02:41 PM
from the anthropomorphism-loves-you dept.
from the anthropomorphism-loves-you dept.
snydeq writes "InfoWorld's Logan Harbaugh follows up his '10 IT Power-Saving Myths Debunked' to argue in favor of using DC power in the datacenter. The practice — viewed as a somewhat crackpot means for reducing wasteful conversions in the datacenter just a few short years ago — has gained traction to the point where server vendors such as HP, IBM, and Sun are making DC power supplies available in their server wares. Meanwhile, Panduit and other companies are working to bring down another barrier for DC to the datacenter: a standardized 400-VDC connector and cabling solution. And with GE working to list 600-VDC circuit breakers with the Underwriters Labs, DC's promise of reduced conversion waste could soon be commonly realized."
Related Stories
[+]
10 IT Power-Saving Myths Debunked 359 comments
snydeq writes "InfoWorld examines 10 power-saving assumptions IT has been operating under in its quest to rein in energy costs vs. the permanent energy crisis. Under scrutiny, most such assumptions wither. From true CPU efficiency, to the life span effect of power-down frequency on servers, to SSD power consumption, to switching to DC in the datacenter, get the facts before setting your IT energy strategy."
This discussion has been archived.
No new comments can be posted.
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
Full
Abbreviated
Hidden
Loading... please wait.
The arguments of olde (Score:5, Funny)
The arguments of olde - don't carry much weight (Score:5, Interesting)
Alternate view: http://cim.pennnet.com/display_article/347089/27/ARTCL/none/none/1/A-powerful-debate:-AC-vs-DC-distribution/ [pennnet.com]
Or, to summarize - if you take a high-efficiency AC system and convert it to 480 volts, downstep to only 240 volts (and all todays' boxes can run either 110 or 220-240), you can get to within 1% of the DC system.
Add to that the savings in materials (1.5" copper wiring? Booster cables for diesels aren't anywhere near that thickness) and there's no real reason to change.
In fact, the biggest saving would probably be if we went from 120v to 240v for everything. One less down-conversion, etc.
Parent
Re:The arguments of olde - don't carry much weight (Score:5, Interesting)
You can achieve substantial savings just by wiring your datacenter for 240V only (in the US). The rest of the world knows this already, but every time I suggest this in the US, people look at me like I have monkeys flying out my nose. Half as many amps == half as many power strips, half as many UPS devices, half as much wire, etc. With the exception of cheap-ass wall wart powered devices, I have not encountered any equipment that was not 240V compatible in the US in years.
Parent
Nonsense (Score:5, Informative)
Half as many amps == half as many power strips, half as many UPS devices, half as much wire, etc.
In the split-single-phase arrangement that is used in the USA, the only difference is whether there's a neutral wire in the conduit. For a given wire gauge you don't get any more power from a 240V circuit, because they're fundamentally the same thing, just one has kind of a "center tap". That copper is a very marginal savings (3 conductors vs 4) when you figure all the labor, conduit, breakers etc that's going to be put in anyway. And if you're dealing with 3-phase it's even less (4 vs 5 conductors).
In a colo environment it would be smarter to run 120 (with shared neutral) so people can use the normal plugs and cables that they have on hand, although in a single-customer datacenter where all your equipment is sure to have modern power supplies, fine, go with 240. But it's not hard to wire 240V outlets as needed (eg for a high density unit like a blade chassis or cisco gear).
You don't use any fewer power strips because you still need a plug per computer regardless of the voltage, and you still need to same amount of UPS equipment because your VA and WH would be the same for a 120 vs 240v UPS of a given price or physical size. It may surprise you that 120V and 240V UPSes generally have the same internals, the only difference is the plugs and cables that they're outfitted with on the back panel. Try measuring the voltage across two hots on different plugs of a "120V" UPS - you'll probably see 240V.
Parent
Re:The arguments of olde - don't carry much weight (Score:4, Informative)
Half as many amps == half as many power strips, half as many UPS devices, half as much wire, etc.
|
This is all utterly and completely false. Number of amps does not affect number of power strips as other posters have proven with math the first computer could have done. (20 == 20 => true)
The number of amps does not affect the number or capacity of UPS devices; it is Watts (commonly referred to as volt-amps) which dictates this, and it remains constant for a specific device no matter the supply voltage.
Your wire savings formula is flawed, unless the 240/415V technique is used as proposed (sort of) by another poster.
As mentioned by another poster, data centers are 3 phase electrical installations. This means there are three wires with alternating voltage in them, and they are all at peak at different times. A wiring technique is used in many installations called "Edison Three Wire" (I have no idea if they use this in data centers, but if they don't, they're stupid) This brings two live wires and one neutral wire to a location requiring two circuits of a given amperage. Let's call it 30A since this requires 10GA wire on a short run from the panel. Two Circuits, three wires. If you remove the ability to use Edison three wire by using 208 V circuits involving no neutral and two live wires, you increase the wire usage by 50%. An up in voltage yields a down in voltage by the same factor. The savings in current is only 42%. 150% of the number of wires times 57% as much ampacity. 17A requires 12GA wire, or run 14GA wire and a 15A circuit breaker which you pray no one trips on a daily basis. (Wow that's reliable) This does not factor in the need by law for each circuit to have a separate ground wire, adding to the number of wires.
10GA wire has a cross sectional area of 10.4 thousand circular mils or kcmil. (mil = 0.001 inch) (1 circular mil is when you have circle radius 0.001 inch.)
12GA wire is 6.53kcmil
14GA wire is 4.11kcmil
Wikipedia on AWG [wikipedia.org]
For arguments sake and the fact that calculations dictated that a circuit now needs 17A down from 30A, we'll use 12GA wire. If you want to argue that it could be 14GA consider the fact that if you cut the available current by 2A you will likely need to increase the number of circuits to an enclosure or other fixture such as cooling or lighting. This will require larger distribution panels, bigger feeder cables, larger conduits, and all around more electrical capacity when dealing with things such as generators and UPS. This will eat into the savings.
Your new wire is 65% as much copper as the old wire. You require 150% as many wires not counting grounding/bonding. Your total mass of copper used is now 97.5% as much as before. A total savings of: precisely dick.
Or, you can double the voltage of every circuit in the data center and leave the electrical network topology the same. This requires new transformers, new distribution equipment, and now run the risk of never being able to provide a customer a 120V circuit for their wall-wart powered device. (I'm assuming that's a transformer block, many of which are now supporting 240V anyway) You could save about 60% as much copper mass, and then spend 10x more replacing all the other equipment which delivers the electricity around the data center, keeping in mind the cost of hiring a certified electrician to install this is tremendous.
Wire is the cheapest piece of equipment in the entire building, and it's the only thing that will be saved in the 240V datacenter, even if you start a brand new building from scratch with this in mind from the first mark on the design plan. Get over it. No one wants to do it.
Perhaps an electrical engineer could come up with some more promising data for converting data centers to 240V up from 120V, however I'm quite certain an EE wouldn't say "WHERE'S MY FISH YOU IDIOT"
Topic Change
Drifting away from t
Parent
Re:The arguments of olde - don't carry much weight (Score:5, Insightful)
Only half of his nose... your server draws a certain number of watts, if you are feeding it 240VAC, it will take 1/2 as many amps as 120VAC (assuming equal efficiency at both voltages, I believe 240VAC-DC conversions are typically marginally more efficient than 120VAC-DC)
P=I^2*R, power loss in your building wiring is based on the resistance of the wiring multiplied by current squared, so dropping current by half will drop your wiring related losses by a factor of 4. Since wiring power loss isn't much of a big deal, a better way you can capitalize on this advantage is by safely dropping your wire diameter by half (to 0.25x cross section) for the same heat/fire potential.
Of course, your boss will fire you when he plugs in a 120VAC coffee pot and it explodes in his face.
Parent
Re:The arguments of olde - don't carry much weight (Score:4, Insightful)
Of course, your boss will fire you when he plugs in a 120VAC coffee pot and it explodes in his face.
Yeah, but you don't take a coffee pot into the datacenter.
Come to think of it, you don't take the boss into the datacenter either...
Parent
Re:The arguments of olde - don't carry much weight (Score:5, Informative)
Parent
Re:The arguments of olde - don't carry much weight (Score:5, Informative)
The concept is actually to go with a European 240/415V system rather than ever using US voltages of 480/277 and 120/208V; you step down from medium voltage directly to the 400V. "Best practices" would be to have an offline or line-interactive UPS.
The biggest gain is actually in the power supplies and not the electrical distribution system. I'm a fan of 600VDC in the data center from an engineering perspective, but there are huge safety issues that need to be resolved to make it viable. (DC arcs don't self extinguish as there is never a zero crossing.)
When discussions were first being done five years or so ago, my theory was that for it to be practical you would need a 3N design rather than today's 2N system, as all work would need to be done on cold busses and you still had to maintain 2N redundancy.
Parent
Re:The arguments of olde (Score:5, Interesting)
Ere... not sure why "Insightful" since Tesla was the one who invented the AC polyphase distribution system, and would probably not approve of using Edison's (not Franklin's?) DC distribution method.
That said, AC power made a lot more sense before the event of solid state power electronics. You can't reasonably convert DC to DC efficiently without using an AC phase via transformer, which was a major hurdle in using DC power. High frequency power supplies can do the job just fine, though.
=Smidge=
Parent
Re:The arguments of olde (Score:5, Interesting)
I think Tesla would be just fine with DC power if he saw what we're using it for today. Back then, there wasn't much stuff that cared which way the current flowed. Lights and electric heaters work fine either way, and motors are more efficient on AC, as is any power source that depends on spinning a generator (almost everything besides solar cells). But once you start throwing diode junctions and electrolytic capacitors into the mix, things change.
Parent
Re:The arguments of olde (Score:5, Informative)
Tesla was not discriminating against DC power in general, he was merely certain that it was AC electricity that was the winner for transporting electricity over long distances, to which Edison objected in favour of DC, but Tesla turned out to be right. To my knowledge, Tes;a never objected scientifically to DC being used in wherever else it was due - such as medium and shorter path interconnects and fine electronics where precise voltages were needed.
Parent
GE working with DC. (Score:5, Funny)
Who would have thought the GE would be a big supporter of DC.
only because (Score:5, Informative)
Switching power supplies have gotten much more efficient in the past few years. Now it makes sense for a standard DC bus to run everything. The telecoms have been doing this for ages.
Parent
Re: (Score:3, Insightful)
Telcos have been doing it at -48VDC for ages. I'm not so sure -400VDC is a good thing. DC voltage doesn't let go, if you grab a wire by accident you'll be toast
-48VDC is safe, -400VDC is scary
Re:GE working with DC. (Score:5, Informative)
Parent
Re:GE working with DC. (Score:4, Insightful)
As far as using DC in the datacenter, my calculations show it just doesn't pay, one or two percent more efficient power use does not justify the large premium DC parts demand today. Part of this is economies of scale and part is market segmentation, DC has historically been used for carrier grade equipment which equipment manufacturers have been able to demand a premium for.
Parent
Battle of the Currents (Score:5, Funny)
I felt a great disturbance in the Force, as if millions of Tesla fanboys suddenly cried out in terror and were suddenly silenced. I fear something terrible has happened.
Re:terrible indeed (Score:5, Informative)
Here's a table:
http://standby.lbl.gov/summary-table.html [lbl.gov]
Doesn't look like they have a breakdown on lcd and plasmas, but rear projection TVs are listed as 6 W when off. Non-DVR set-top boxes for satellites are about 15.5W, and DVR types are almost 30W when not even recording anything.
Parent
600VDC is not chicken soup for the soul (Score:5, Insightful)
Suggestion for the DC power supply designers: have a heart and build GFCI into the spec.
What about a _home_ standard for DC power? (Score:5, Insightful)
I don't run a datacenter, but I sure would like to get rid of the power bricks that all small electronic appliances seem to come with these days!
Re: (Score:3, Interesting)
Agreed. My PC and media installations are plagued by a plethora of these heat-generating devices, as I add on printers, ethernet devices, networked disks, extra storage, converters, encoders, decoders, and the like. I had to learn to include plans for a well-ventilated place for these things.
Also, it's an inherently good idea for power savings. Power supply efficiency can go way up when both a) total power goes up and b) the supply can be designed for a constant load (which would be the case for a large
Re: (Score:3, Interesting)
I don't run a datacenter, but I sure would like to get rid of the power bricks that all small electronic appliances seem to come with these days!
probably because these 'wall-warts' are linear converters - seldom better than 40% eff.
As more stuff conforms to the ENERGY INDEPENDENCE AND SECURITY ACT OF 2007, these will become much less of an issue.
This is snake oil (Score:5, Informative)
> I don't run a datacenter, but I sure would like to get rid of the power bricks...
DC vs AC wouldn't help you rid yourself of power bricks. No more than it can help a datacenter get rid of power supplies in each server. Telco equipment runs on 48 volts not to save electricity but because of the way telephone exchanges are built. Telephones don't go down, period. So how do they accomplish this miracle? Huge battery banks. Back in the day a DC-AC conversion system large enough to run a whole switch plus drive every telephone would have been all but impossible. So they just ran everything directly from the batteries and used the mains to charge the batteries.
This DC in the datacenter thing is just a green craze that will pass. It is pure unadulterated snake oil. Go reread the summary. They ain't even doing the smart thing and adopting the telco 48V standard. Does anything in a server run on 48V? No. Does anything in a server run on the 400V they are proposing? No. So a DC-DC conversion will be needed, i.e. a switch mode power supply. Guess what is in a current server? A switch mode power supply. Current PC power supplies are available with efficiencies over 90% without buying too far off the mainstream. I seriously doubt these DC powered supplies will be much better and in the end that is the ONLY number that matters. Except these DC installations have to factor in the power loss from the big AC-DC conversion and worry about redundency, backup power, etc.
Parent
Re:This is snake oil (Score:5, Informative)
Switch mode supplies that run off DC input don't require a big high voltage input capacitor. They also don't require complicated PFC circuitry. Basically, a modern AC-DC SMPS has an input boost converter that goes to ~380-400VDC, and then a forward or flyback converter that turns that into usable voltages. This is required to get power factor correction, which is required for high efficiency on a large system. This system moves the first half of that outside the computer into one large device. Running the large converter off three-phase power makes it mildly more complex but removes the bulk capacitors. Between that and the fact that there is only one of these converters, it's a lot cheaper. Also, power electronics generally get more efficient as they get larger, for a variety of reasons; this takes advantage of that.
The major economic reason to run off 48VDC instead of 400VDC is that some gear already exists thanks to the telcos; the major reason not to is thicker more expensive wiring. Which one wins depends on the size of the market, and it sounds like the market is big enough that the 400VDC probably wins.
If you really wanted to, you could push the AC-DC efficiency higher with more expensive electronics -- but centralizing it is cheaper, so why bother?
Parent
Re:What about a _home_ standard for DC power? (Score:5, Interesting)
There is one already: USB power. Fairly low current, but a host of consumer devices from bluetooth headsets to GPS devices to iPods use it as their standard charging source.
It's a little awkward because there are more pins than ought to be strictly necessary, but it's a relatively reasonable compromise over the former solution of no common standard at all.
Parent
In the good old days (Score:4, Informative)
In the 80's we built custom interfaces for large computers using wire-wrap Standard Logic Inc. wiring modules. The planes of wiring were assembled into rackmount chassis which were fed DC power via a vertical bus-bar system in the rack. The busbars were about .5 X 1 inch solid copper, insulated by shrink tubing with holes cut for the threaded holes in the busbar. The power supplies were rackmount 100 or 200 A Lambda supplies providing either 5 volts or 12 Volts. It was occasionally a pain to be called into the computer center in the middle of the night to replace one of those heavy power supplies - at least they were at the bottom of the rack.
OZ
WTF? (Score:5, Funny)
Re:WTF? (Score:5, Informative)
The article can basically be summed up as follows:
Though there are more transmission losses with DC than with AC, if your DC->AC conversion can be done with an outdoor-rated supply, you save more in cooling by doing the conversion outdoors than you'd lose in transmission losses.
Parent
Re:WTF? (Score:5, Informative)
DC power lost the "current wars" because we didn't have solid state transformers capable of doing voltage step up/down like we did with AC back in the day (simple wound transformers).
These days even the cost of really high power DC transformers (>500,000 volts) is offset by more efficient transmission and a number of notable long-distance power lines are actually DC for that reason (lower losses offset cost of transformers).
By stepping up the voltage, such as to 48v, you can significantly lower the losses, shrink required conductor sizes, make the circuit breakers cheaper, and still derive the same benefits (48v->12v->5v->3.3v DC transformers are actually fairly cheap, unlike their high-power cousins).
Why do you think some car makers are switching to 48v DC on-board power and 48v batteries? You can greatly increase efficiency and lower weight since so many devices are electrical on modern cars.
Parent
Re:WTF? (Score:5, Informative)
In fact AC loses slightly more at a given voltage, up to a lot more for really long wires.
Whiskey. Tango. Foxtrot. Line losses are based on current not voltage. And with AC you can convert current and voltage with a transformer with a very high Q. That's why AC (Tesla) beat DC (Edison) at the turn of the century for power distribution. Also, direct current generates more heat than alternating current. -_-
Parent
Re:WTF? (Score:4, Informative)
No WTF.
At a given voltage/amperage, DC will lose less power per mile than AC. However, AC transformation equipment is cheaper/more efficient than DC.
At a couple hundred miles, DC becomes the more cost effective solution for a high power run.
Also, direct current generates more heat than alternating current
Not at the same wattage.
Parent
Old hat in the telco world (Score:5, Interesting)
Telco gear tends to be 48VDC all over the place. It just works. Speaking as a guy working at a telco in the IT department, I'm hugely in favor of moving to 48VDC servers.
Why not 12V, 6v or 3.3v, etc? (Score:4, Interesting)
I'm not an EE. But back during the dotboom I thought it would make sense to have a big ups in the data center that output voltages that mother boards expected as input. I almost thought of rigging my own experiment using laptops as servers and feeding them all 12vdc directly from the UPS battery pack.
Ok rip it apart guys, why is wrong with that plan?
Re:Why not 12V, 6v or 3.3v, etc? (Score:5, Informative)
Power loss over distance. 12 volts loses four times as much energy in one foot of travel as 24 volt transmission does. Telecom gear, for example, runs on 48 volt DC. For the few feet of travel in your laptop, 12 volts is fine. Crossing a room at 12 volts, you'd get too much voltage drop.
See http://en.wikipedia.org/wiki/Electric_power_transmission#Bulk_power_transmission [wikipedia.org]
Transmission efficiency is improved by increasing the voltage using a step-up transformer, which reduces the current in the conductors, while keeping the power transmitted nearly equal to the power input. The reduced current flowing through the conductor reduces the losses in the conductor and since, according to Joule's Law, the losses are proportional to the square of the current, halving the current makes the transmission loss one quarter the original value.
Parent
48VDC pros/cons (IMHO) (Score:5, Interesting)
Pros:
No power supply needed for each machine. This removes a major point of failure. Instead, one would need to just step down voltages to the 5 and 12 volt rails. This also helps with cooling because the room AC/DC converter can be cooled with a dedicated system, either liquid, or part of the HVAC system.
Cons:
48 VDC needs a dedicated connector with a high plug/unplug cycle rating that people know is 48 volts and 48 volts only. It sucks when you have to manually wire it up, because this takes time and there is always the risk of getting zapped if you don't throw the right circuit breaker (or pull the right fuse) on a telco rack where 48V is in use.
Because there is only one 48VDC power supply for a room, it has to be held up to a lot more rigorous standards than average mains current. It has to not just provide 48VDC, but provide it under extremely heavy load without the voltage dropping by much.
Maybe 48 volts would be a new computer standard. The key is not having to wire it up manually like some stereo speakers, but giving it a dedicated, foolproof, power connector that Joe Twelvepack who is slurping down his seventh can of Bud Light can easily and reliably plug and unplug while staggering around in the back of the server room until his shift ends.
Re:48VDC pros/cons (IMHO) (Score:5, Insightful)
Another pro:
A UPS would consist of nothing more than a battery charger and 48V battery.
Parent
Re:48VDC pros/cons (IMHO) (Score:5, Interesting)
Parent
Re: (Score:3, Interesting)
The whole 48v DC thing sounds good to me (I don't run a data center though, or anything like it).
That said the article discusses (and I've seen it said elsewhere) the large copper bars used for wires in this kind of setup, and how they will lose more power between the wall and the rack than AC.
I can see the appeal of going TOTALLY 48v, but why not run AC to the racks, and just have a large converter for every 2 or three that provides the full DC power and backup for those three racks? You're still avoidin
Edison vs Tesla (Score:4, Insightful)
One can't help but reflect upon these two and their stubborn support of DC and AC respectively. Edison created a circus atmosphere demonstrating the dangers of AC. He electrocuted dogs & other animals and even participated in the design of the electric chair to prove his point.
Edison's financial ambition was part of the problem, and his inability to understand AC, but mostly it seems to have been an emotional attachment to DC.
Let's hope that in our time emotion and personal gain have no part in such decisions.
Already on board (Score:5, Funny)
a standardized 400-VDC connector and cabling solution
I set this kind of system up myself and it works great, assuming you need a lot of cores. I strung together 296 Intel Core 2 Duo chips in series accross the 400VDC supply, so each one gets the specified 1.35 volts. If I want to overclock, I just take a set of alligator clips and shunt across a few dozen of the chips, and it boosts the voltage to the remaining CPUs.
The only problem is that with so many chips, I get occasional failures, just like I do with my old Christmas lights. Then I have to try shunting around each of the CPUs by trial and error until I isolate the burnt out one before I can get my cluster running again. Oh yeah, I also have to be really careful to keep any peripherals I plug in away from each other and/or grounded objects.
Already seen on UPS units (Score:4, Informative)
(Since to use a UPS without DC means converting battery's DC, sending it to the PSU in AC where it's converted back again.)
IIRC there was a UPS study on DC power recently (Score:4, Interesting)
That concluded that using the european system of 230/400 3 phase AC for distribution splitting out to 230V single phase AC near the point of use was almost as efficiant as a 400V DC system and far cheaper and easier to deploy. Your servers existing power supplies can almost certinaly handle 230V without any problems (changing a switch may be required on crappier models)
BTW in many cases there are often huge savings to be made without changing your infrastructure just by using better PSUs, cheapasss PSUs are both inefficiant and unreliable.
Commercial DC Data Center solutions (Score:4, Informative)
What's old is new again? (Score:4, Informative)
The reason Tesla/Westinghouse won the current wars with Edison because there wasn't any good way to step DC voltages up or down. You can't transmit power very far at 110 volts. AC allowed the use of inexpensive and transformers to step voltage down at the customer site and transmit at high voltage over long distance.
Today solid state converters do allow the step up / down of DC voltage, and very high voltage DC can be sent over long distances with less loss than the same AC at the same voltage. At least one power company is looking at using DC transmission lines over long distance.
AC power still makes more sense for consumer and most industrial use, but for transmission and delivery of power in bulk DC seems to be making a comeback.
Re:And Nikola Tesla.... (Score:5, Funny)
rolls over in his grave.
If that's because of DC, does it mean he's fitted with a brushless commutator?
Parent
Re: (Score:3, Funny)
Re:Why wasn't this tagged 'edison v. tesla'? (Score:4, Funny)
Assuming that you don't modulate the phase variance of the deflector dish, of course.
Parent
Re:Why wasn't this tagged 'edison v. tesla'? (Score:4, Informative)
I call not true!
The body is exceptionally good at accomidating a stable force acting on the system. What causes most electricution deaths are the sudden change in voltages throwing the heart out of rhythm or scrambling the brain log enough for the person to die.
The actual physical damage of electricution is usually very minor (first or second degree burns through the path of the current). The alternating nature of AC makes it much more likley to mess up the heart and brain. 120 chances a second. DC only has one chance.
Now DC will cause greater BURNS because the constant voltage at the same power can generate more heat, but the burns are not what kill you.
Neither article you cited mentioned DC vs AC. Almost every mention of current related it to HOUSEHOLD current which suggests AC.
Finally, the blood cannot be "charged." It is a fluid with some conducting ability since it is full of various ions. Any charge it does accumulate would almost immediatly ground out to the rest of the body and from there to the earth.
If you want to make dramatic claims please provide plenty of citations
Parent
Re:Why wasn't this tagged 'edison v. tesla'? (Score:5, Insightful)
However, in these links there was no reference to this at all.
I don't think there is any truth to this.
Parent