Slashdot Mirror
AC and DC Battle For Data Center Efficiency Crown
jfruh writes "AC beat DC in the War of the Currents that raged in the late 19th century, which means that most modern data centers today run on AC power. But as cloud computing demands and rising energy prices force providers to squeeze every ounce of efficiency out of their data centers, DC is getting another look."
Yeah, I know the article only mentions one prior Current War, but that doesn't fit the template correctly.
AC is better than DC for transporting electricity because you can convert between voltages with just a transformer. But in a data centre, when all the equipment will be powered by the same voltage, it makes sense to use one good efficient power supply for multiple computers, so that all the components don't have to be duplicated for each computer.
It should be noted that this battle is on a different battlefield than the one in the 1880s. There the battlefield was distance and AC eliminated the need to have a power plant every few miles. In addition it takes lots of copper to carry lots of current any distance. Now a data center still is AC to the PDU or power switch board where it may well be converted. The outside power will still be AC and likley the generator set will be AC as an AC generator is lower maintenance than a DC one. On a related topic locomotives are switching to AC from DC as the weight of the motors and the generator set is less. It is perhaps more appropriate to say that the choice of which kind of current depends on the application. Due to better rectifiers and inverters very long distance power transmission is better done with DC, to avoid the capacitance problems with AC on long high voltage lines (The Dalles OR to LA for example)
which means that most modern data centers today run on AC power
Only if you ignore all telecom equipment which have run on -48VDC for decades. True, they're not really 'data centers' but it's not like they don't use massive amounts of electricity.
If you use a high voltage, you need power converters in each computer. If you use low voltage, you need thick, expensive, hard-to-run wire.
Most installations will be much better off sticking to conventional AC power.
most telephone exchange and related transmission hubs use DC 12, 24 and 48VDC are standard. This isn't anything new, and data centers have always been space and power inefficient, it's the nature of the beast, and method of construction.
There was an unknown error in the submission.
When talking about DC power distribution in data centers, people usually assume the DC will be at a low voltage, like 12, 24 or 48V. With such low voltages, the currents must be huge to supply the required power, which translates into large losses, or very big cables and busbars.
High voltage DC, on the other hand, really makes sense; instead of using power supplies with internal rectifiers and power factor correction, with the latter usually causing a significant loss in efficiency, one could build a big, central power factor correction device, which is probably more cost efficient en possible more cost effective than having a separate one in every single computer power supply.
AC makes sense when you need to use old style transformers (big iron cores, at 50 or 60Hz), while switch-mode power supplies would work a lot better with a high-voltage DC input. An added bonus is that the RMS current in the cables is smaller with high-voltage DC, reducing the losses even further (although the gain is probably rather small)
it has nothing to do with AC versus DC, but when and where the conversions are done.
In an old-style setup, you use high voltage AC for long distances and convert to DC just once, near the load.
In a medium-new design, you convert the 120 or 220 VAC to DC by rectification at each PC, then use a switching-mode power supply to down-convert that to the required DC levels.
In an even newer design, you notice that the rectifiers are semi-redundant, you can factor that out and have one or a few larger rectifiers and distribute medium-voltage DC.
You will still need switching-mode regulators at the card or CPU level.
Not a very big difference. Factoring out the rectifiers lets you use slightly more efficient synchronous rectifiers. The money you save there you can apply to buying slightly more efficient switching regulators.
There was an article about using 380 volts a couple weeks ago on /. in the data center.
Having DC brings some benefits, mainly just needing to step down voltage and not have to rectify it smoothly with capacitors to even out the output current.
However, there are some downsides:
1: AC power supplies in devices tend to be more tolerant of power fluctuations. An all DC shop might completely be halted by a power surge/spike that wouldn't bother a data center on AC.
2: DC sparks a lot when connecting/disconnecting. AC has plenty of zero-crossings a second (120 or so), so it won't make the fireworks show when plugging/unplugging. This makes switches rated for DC a lot more expensive than AC.
3: There is no such thing as a NEMA 380VDC connector. So, either items would have to be wired up to a bus bar similar to how 48VDC telco stuff gets, or it will end up like 12VDC with at least 5+ connectors (direct wires, cig lighter, airplane, marine connector, male/female combined connector, motorcycle accessory connector, banana plugs.)
4: Safety. 12 VDC shocks are annoying; a shock from 380VDC will be fatal, especially because of DC's tendency to get muscles to "lock". (This is why stun fences uses AC, while kill electric fences use DC so they can keep the target locked on the wires long enough to get the amps across the heart.)
5: Issues with wire length. AC, it isn't hard to use a transformer to deal with voltage drop. DC, that will be a lot harder.
All and all, 380VDC seems like a solution in search for a problem. We really don't need another standard. Heck, just pointing out 120VAC in the US means I have to doublecheck if I'm dealing with 15 amps, 20 amps, 30 amps, or 50 amps, and the locking versions of each, which means six plug types and minimum wire gauges.
Are data centers finally ready for DC power?
http://hardware.slashdot.org/story/11/11/30/213247/are-data-centers-finally-ready-for-dc-power
Are they? ARE THEY?
Why not send prototachyons through the main deflector array?
AC, DC, it does not make a difference any more. Yes, you have to rectify AC before it powers a computer, but the rectification costs less than 1% of the energy. Power factor compensation can be more costly, but it could be avoided by going to a 3 phase rectifier. There are also serious distribution advantages in 3 phase electricity, but it is not used because of the extra complexity, despite being cheap.
DC distribution is expensive, and 1% gain is just not enough to pay for it. Once we have intelligent grids, the situation may be different, but for now there is just no business case.
What if you want to electrocute an elephant?
Give me Classic Slashdot or give me death!
Standard -48VDC current distribution requires four times the current as 208V AC distribution for the same amount of power. Have you seen DC cabling at data centers that use it? If we're going to start using DC in data centers we need to come up with a higher voltage standard, otherwise we're going to spend all the savings on more copper (which is expensive!) to carry those extra amps.
there is something to be said for higher voltages for transport. line loss is I^2R where R is the resistance of the wire. if you need a LOT of power moving it high voltage to where you need it is best. is DC better than AC? i wouldnt say that is true. low voltage DC is going to be a killer to distribute efficiently given the wire losses.
you forgot the lightning bolt
If it was good enough for Edison, it's good enough for me!
Chas - The one, the only.
THANK GOD!!!
Thank god, I'm so tired of stray dogs wandering into my datamacenter and getting electromacuted by deadly AC currents!
Peta's really on our asses about that.
the preceding comment is my own and in no way reflects the opinion of the Joint Chiefs of Staff
Articles main source says modern AC and theoretical DC are about the same. By the way, he has a product to sell...
My mind works like lightning. One brilliant flash and it is gone.
In 2005 we started looking at blade chassis and tested a rack of HP BL series blades.
That system came with a 48v DC power enclosure with 6 hot swap power supplies. It sat in the bottom of the rack and had a buss bar system to feed every chassis in the rack.
As others have stated.. 48v is a long standing standard for telecom power.
If you think it's expensive to hire a professional to do the job, wait until you hire an amateur. --Red Adair
The problem is it wasn't back when the grid was being made. There was no good, easy, efficient way to convert DC voltage. Now, not so hard.
I was shakin' at the knees Could I come again please? Yeah the ladies were too kind You've been - thunderstruck, thunderstruck Yeah yeah yeah, thunderstruck
... convert that AC to DC at a "blade rack". That would be a rack designed to take blades. But the blades would be a mix of
This will safely segment the power, leaving the DC busses limited to the amperage needed for one rack ... or even partial rack. It also has the flexibility of balancing power conversion vs. 1st tier power backup (at the point of use). Increasing the backup times to a couple minutes allows slow start generators, which are more reliable.
I would run 416/240 three phase everywhere in the data center (even in North America ... transformers for this are readily available). Where equipment isn't on the DC system, run it on 240VLN. The AC/DC converters might run on 240VLN or 416VLL. In countries with 400/230 or 380/220, just use it that way direct.
AC is safer due to the zero crossing. Circuit breakers can break a lot more power (usually 5x the voltage) with the advantage of AC, as compared to DC. A 380VDC breaker for a rack would be HUGE, especially if it has to handle a data center level of fault current.
now we need to go OSS in diesel cars
The problem is it wasn't back when the grid was being made. There was no good, easy, efficient way to convert DC voltage. Now, not so hard.
AC won over DC primarily because of the distance it can be transmitted.
whichever one gets Dirty Deeds Done Dirt Cheap
General Relativity: Space-time tells matter where to go; Matter tells space-time what shape to be.
what about PSU with build in UPS hook ups? so you can get rid of the AC to DC to AC to DC part and make it just AC to DC? at each system?
You are getting that wrong. DC can be transmitted farther than AC. DC has only resistive losses, while AC also has capacitive and inductive ones.
I'd sumarize it as the following:
DC is slighlty (just slightly) better for transmitting;
AC was easier to convert from one tension to the other (currently, we have the oposite situation);
AC is better to use on motors (it was much better, now it is just slightly better);
AC is easier to generate (it was much better, now it is just slightly better - except on photovoltaics);
AC is easier on the connectors (hight current DC connectors are a hell to maintain)
It is easy to see why AC won. I bet AC would win again just because of the connectors and generators, after all, converting it to DC is relatively cheap. The only problem is the low frequencies we currently use, it would be better to increase them a lot now that we have better materials.
Rethinking email
I don't want to live on this planet anymore.
What was missed in the 100+ years that electrictity has been in use was the adaptability of use on a broad scale. N. Tesla was right; unfortunately; the 'powers that be' were interested in profit. The 'Wardenclyffe Tower' Experiment should have been a continued. The realization that ' Earth is one big battery' would have changed our perception of electricity, for ever.
AC is what is normal for power companies to provide in the US.
DC is what all of the phone company equipment runs off of in the mini-C.O.s, though.
So, I'd say that using DC to power electronics is hardly a novel concept.
I worked there for 7 years. I'm not going to get into specifics but I will say:
Verari tried to take advantage of the efficiency gains in DC with exotic power supplies etc... And that company went the way of the dodo bird after trying to force 800V, 48V, and 12V DC power distribution systems in customer data centers. The fact is, everything already out there (switches, routers, servers, etc) uses AC-DC power supplies in each unit and it works in 99% of power outlets with pretty good uptime. The added complexity of running DC infrastructure isn't worth the efficiency gains (which on paper sound like a lot but theory rarely translates to reality the way we think it will), and when one DC rectifier burns up and takes down a hundred servers (vs 1 server with an AC-DC supply), customers aren't happy. Between the uptime issues and employee safety concerns (high amperage DC power is more dangerous than AC for a variety of reasons) it's also a liability nightmare
Again, I don't feel like getting into specifics but modern datacenters != underground telco installations and DC power distribution has a LOT of challenges that are often overlooked when marketing types start squawking about efficiency gains.
Uh the article the post links to supports AC more than DC in case no one noticed. The article is about DC being hyped beyond the facts and that AC is claimed to be just as good. Sort of reverses the whole discussion here making it AD, alternating discussion. Edison gets the carbonite filament..
Some drink at the fountain of knowledge. Others just gargle.
You need low frequencies for the grid, else the capacitive coupling to ground will limit the maximum AC cable length yet further and increase eddie current losses in the wiring..
Semiconductor technology has not advanced far enough to justify replacing HV transformers and overhead lines except for very long cable runs, underwater cable runs or connecting asynchronous grids.
AC is better to use on motors, however variable AC is better still - and this can be generated slightly easier with DC (although matrix converters are good, the current harmonics causes nasty problems...)
AC is better for generators, however this is at low voltage so power electronics works well.
Running a data center on DC makes a lot of sense. Even just using the same power supplies and wiring will save a huge amount of power - the rectifiers and the line filters could just be bypassed and the AC-DC conversion could be performed by multiple redundant 12 pulse rectifiers resulting in nice sinusoidal input currents for lower transformer losses.
However have you ever tried to fuse DC. You can get >50A breakers relatively cheaply but anything bigger will cost thousands...
you would know:
- Adding batteries to a central redundant DC power supply makes a wonderful online UPS.
- DC/DC converters are small and efficient.
- PC/Server AC power supplies are not so efficient and not so small.
- AC power supply efficiency is not linear (power usage vs. max. power offered).
A central DC power supply will have a shared high power usage causing a higher efficiency than having an AC power supply for each server.
And for SPOF people: A & B paths are standard for DC power.
In what should come as little surprise to slashdotters anywhere, the telecom equipment switch to -48VDC was driven by a myth propagated by PHB's who were measuring the wrong thing http://www.pipelinepub.com/0407/pdf/Article%204_Carrier%20Grade_LTC.pdf
Take some advice http://www.apcmedia.com/salestools/SADE-5TNRLG_R6_EN.pdf from an engineer (not me) with 19 patents related to high-efficiency and high-density data center power and cooling infrastructure and read up on why so much of the non-US data centers use 400/230V AC because they are (a) nearly 6% more efficient than 48V DC systems and around 1% less efficient than 380V DC. Which helps explain why much more equipment is available for 230V AC than for 380V DC.
I am strapping the electrodes to the elephant now...
Must... unimagine... Tesla... frenching... Edison...
The effort to gain acceptance for DC distribution in data centers is being helped by a series of investments by ABB, and the growth of the EMerge Alliance, which is trying to unify DC proponents around a 380V standard. The challenge for DC is that customers don't ask for it, meaning multi-tenant facilities aren't likely to offer it. Also, Schneider says it is "not aware of any data centers moving off of their established, traditional power distribution to DC." In fact, NTT has at least five DC data centers in Japan, and ABB is backing a DC distribution project at a Swiss hosting company. In the US, there are numerous sites testing DC power, which is widely used in telecom infrastructure.
and you can bet your ass that DC wont make its way into any of them anytime soon. Transmission line efficiency is the same for DC as it is for AC. Yes AC can be transmitted at higher voltages and easily converted reducing the current carried by the transmission, but power is the product of voltage and current and thus the loss is still there regardless - you're just using more voltage and less current. The only real benefit of using a higher voltage is that the transmission line can be smaller.
Not mentioned yet that I see is installation cost. The biggest cost of building a data center is electrical. The current standard is 3-phase 480vac power. 3 hot wires and a shared neutral. In a DC system the hots cant share a neutral. At face value you just increased the cost of wire and bus-duct by 50%. And believe you me, there are tons of it in a data center. If you've never seen one you would not believe how much of that there is in a data center. On a balanced load you really only have to contend with wire resistance to the equipment, and not coming back on the neutral. That's not the case in a DC system. In a DC system, your wire distance is effectively doubled over a balanced 3-phase system, and the DC system needs 50% more to wire to boot.
http://www.younow.com/shows
Back when the current wars were happening, there was no good way to convert DC voltages. Edison's model called for lots of local small powerplants to deal with that. AC was easy to convert using transformers. Now DC voltage conversion is easy. Thyristors do the trick nicely.
Because it was hard to convert voltages, you couldn't do HVDC runs unless you wanted it in the home as well.
To me it seems that AC wins. As long as individual devices have some form of usefull state buffering until the DC/AC inverters kick in to generate more AC. Maybe a mother board could be made where only ram is batery backed up. With huge amounts of ram huge history records could be used to effectively snapshot and propel the computer beyond a power glitch.
Is it just me, or do these 'DC in the data centre' articles seem to get posted to /. every few months?
'He who has to break a thing to find out what it is, has left the path of wisdom.' -- Gandalf to Saruman
All DC in a large data centre and you'll need whopping thick bits of copper and equipment that would look at home next to a 500MW generator.
A rack basis is one thing, and can work well. A big roomfull is another.
... thunderstruck! Yeah yeah yeah THUNDERSTRUCK!
Telecom already acknowledges DC as the victor. It's about time the datacenter people also recognize the efficiencies of DC power in the datacenter.
Kriston
I thought this thread was about http://en.wikipedia.org/wiki/Powerage?
Electrocute an elephant with AC: http://en.wikipedia.org/wiki/Topsy_(elephant)
Occasional designer and frequent user of converters here. While DC/DC converters are lighter and smaller, they're not necessarily cheaper, esp in large sizes. They also are not necessarily as efficient for a given cost and size.
Note that most converters have a transformer inside. They have a DC/AC inverter a transformer and a AC/DC rectifier and filter. Yes, you can do it with just an inductor, but since you've got to have magnetics anyway, adding a winding doesn't change much and gives you a lot more design flexibility.
They are MUCH more complex than some wire on an iron core. Even a simple DC/DC converter has dozens of parts in it.
Conventional transformers are FAR more robust than DC/DC converters. Drop them off a truck? have lightning hit near by? DC/DC converters have all those semiconductor devices in them that either need protection (more components) or will fail in a transient overvoltage situation.
That said, there are applications where DC/DC devices work really well. Fluorescent light ballasts are an example. They were done with a transformer with lots of leakage inductance and designed to very low cost, since the market wasn't efficiency sensitive (as long as it didn't actually catch fire, 65-75% efficiency was ok, because poor efficiency ballast + fluorescent tube was SO much better than what it replaced: tungsten filaments). Now, though, that everyone has fluorescents (or other discharge lamps like metal halide, etc.) there is an economic incentive to go to more efficient designs, and it turns out that driving a highly nonlinear discharge load which needs a voltage spike to start it is something that is hard to do with conventional line frequency magnetics.
Motor speed controls are another area. most motors have good efficiency only at one speed and power output, but most loads actually have variable speed needs (fans, pumps, etc.). With a solid state drive, you can get the overall system to be more efficient (run motor at optimum current and voltage for load, so motor losses are greatly lowered, more than you spend on the variable frequency/volltage drive).
The losses in a modern transformer are very low (1% for large power distribution).. a 1 MVA transformer that was 1% loss would dissipate 10kW of heat.. that adds up. (a DC/DC converter that does 95% is doing VERY well. 85-90% is more typical)
Nikola Tesla VS Thomas Edison
FIGHT!
You are getting that wrong. DC can be transmitted farther than AC. DC has only resistive losses, while AC also has capacitive and inductive ones.
That is perhaps the single most ignorant (yet modded up) thing I have ever heard of the AC-DC debate. DC cannot come close to multi-phase transmission of electricity, it does NOT travel further than normal AC when the frequencies are properly tuned to the network of wires and the power levels being utilized and even in relatively short distances AC is better than DC for it's ability to transmit. Hell, you can't even get a DC line running 12+ miles because the thing will lose so much power in ohmic heating it becomes prohibitively expensive to power the line. The ONLY time DC can be transmitted further than AC for a given level of power (and this doesn't even work in many situations) is to have voltage on the order of millions of volts - then the effects leading to it's transmission will typically manage a little better over short durations of time. A simple way to look at it is this: With DC, you are actually moving electrons around in a loop, you get them from the power supply. With AC, you are very unlikely to ever encounter an electron in the wires of your home that came from the power company (save for extremely bad timing on the oscillators driving the system) - AC just pushes and pulls, relying on the pressure of the electrons against one another to transmit them without the extra pushes. I'm shocked you didn't get -5 Troll for that insanely incorrect post.