Slashdot Mirror
Data Centers And DC Power
mstansberry writes "In the final article in a series on the price of power in the data center, IT pros weigh the pros and cons of direct current-powered servers. A limited number of companies make servers with the power supplies removed with DC power distributed to multiple machines from a single unit. It saves power by skipping an extra conversion from alternating current (AC). Telcos have been using this method for years, but some data center pros are leery of taking on the new systems. It's not something people are familiar with and if they break down, you have to hire a specialized engineer to come fix them. But if they're saving even half of what they're reported to save on the electric bill, companies could afford to hire the engineers." We've reported on previous articles in the series.
Does big iron still use 3-phase power?
Tesla, you're fired. --Thomas Edison
try { do() || do_not(); } catch (JediException err) { yoda(err); }
I'd be leery if you wanted me to go homopolar in my center.
I've always wondered (from a non-technical point of view) whether there was a benefit in having our homes wired up with two sockets (or maybe a 5 pin mains plug) giving standard AC voltage and a low-current DC voltage as well (12V?). So many devices only need low voltage, wouldn't we all benefit in having a power system in our houses in this way?
Jolyon
Please read my Canon EOS tech blog at http://www.everyothershot.com
ac goes into data centres, systems run on dc. Either it gets distributed to each computer as ac and converted in a medium-sized box in the back of each system, or it gets converted in one big box and distributed to the systems as dc.
The question is of the efficiency saving of doing all the converting in a big box against the efficiency loss of piping it around the data centre as dc, and wether you get a large total net saving (which I suspect that you do, since even inside the data centre, it's not going far)
FGD 135
trouble shooting and correcting DC power is simpler than working with linear power supplies. Unfamiliarity is the problem, not the technology.
I wish homes were wired for DC. I think we'd save a lot of power doing the conversion in one place per household/street, rather than using a separate transformer for each device. Plus you'd wouldn't have to waste time trying to find your devices transformer, or waste space when packing the device when you go on holiday.
I've often pondered why we don't have a central AC->DC power conversion in our home and have DC outlets. Most my electronics run on DC power and they all have some kind of bulky converter. Seems very inefficient.
I wonder how much energy a household could save if we were all using DC outlets instead of AC.
Maybe im wrong but I would think most engineers would be able to learn it pretty easy if they don't know already.
Any truly serious data centre would already have at least several power engineers on their staff.
If such a data centre is just now considering bringing such people in, then they have serious operational problems. They're not getting professionals in to do the jobs that professionals must do.
Cyric Zndovzny at your service.
There is no standard Telco DC connector. This can cause headaches since you need special connectors and a AC-DC box for every type of DC appliance you have. Most every AC appliance uses one of a few types of industry standard AC connectors.
What we really need are pluggable racks. i.e. Move all the hardware necessary to support the blades (power supply, network switch, cooling, KVM interface, etc.) right into the rack case, then design a common interface to plug the blades into. Then an admin only needs to plug in the new server and run with it. No need to mess with tonnes of wires.
Wait. Did I just reinvent a 64 way Sun server? Imagine that.
Javascript + Nintendo DSi = DSiCade
I'd love to see rapid advances in DC power distribution systems. Other areas would benefit, too. Consider alone the reduction in RF noise in electronic systems that would be achieved by eliminating all those transformers and switching power supplies. Granted, DC/DC converters still use switching power supplies - but there lies my wish for advancement.
This sig is a test. If this had been an actual sig, you would be reading something quite a bit wittier than this now.
No, it saves power by having one large, very efficient power supply do the conversion to DC rather than several dozen small ones. Either way, there is still only one conversion: 110AC -> 12/-12/5vDC.
If a job's not worth doing, it's not worth doing right.
Don't forget that DC cables must be much heavier gauge than AC. The savings in copper alone will pay for those little cheap AC/DC transformers.
Also, who wants to run two cables to every socket block?
I don't see what the problem is. Each box would be built exactly the same as before, except it wouldn't have a power supply unit - just a straight wire from the power socket. And the central power supply is just a big AC/DC converter - most EEs should already be familiar with those. There's not really any new technology here, it's just slightly re-arranged.
So for some reason, they convert AC line voltage to DC, then back to AC, then back to DC again in the power supplies? That makes no sense. Transformers go from one AC voltage to another (2k to 120). What's the DC step in the middle for?
And people keep talking about the added efficiency of a DC wall outlet, but that conversion isn't free. That's why those adapters you plug into the wall are always warm, even if you're not using them. You'd be using power just to keep the voltage available.
The only way I can see for them to save power with a distribution center is if they skip the transformer, because that takes power, too. But that's not what the guy in the article said. One of us is confused.
Why do you need the batteries? Many new solar panel controllers have built in grid-tie capability. When your panels make more juice than you use you get a credit because you put power INTO the grid. Very cool, only problem is the panels cost too much.
If you want that to work, we need to get a major advance in solar panel tech. The grid-tie stuff is here already.
If main power fails in an AC data center, the UPS systems need to take the DC from the batteries and convert to AC, then distribute, then each machine needs to convert back to DC. That's terribly wasteful, since neither of these conversions is anywhere near perfect efficiency. In a DC data center, the UPS systems are just the batteries, so they can hold much longer.
Now, of course there's a counter to this; the problem is that batteries don't handle the load very long, and you typically need to switch to a generator, and typical mechanical generators inherently provide AC rather than DC (because it is mechanically more efficient).
The conversion from and to the UPS is saved. Every watt that is lost in the conversions is converted into heat. That heat has to be cooled of by cooling systems using also an enormoes amount of energy.
By saving 20% on the conversion you also save 20% on the cooling. But also the power can now run in a different room where the temperature condition might be less demanding, so even more cooling might be saved.
"His department conducted a study that said 80% of those servers were running at 5% to 15% utilization"
Why such low utilization ?
Any other industry would scrap 80% of that equipment to save costs and power.
I started working with IP in a small ISP. We were bought by a loal Telco and over the years have got used to having all our routers and switches running on DC current.
One thing telco companies do well is DC power, they have alot of skill in providing multiple DC feeds from DC power systems, with battery backup and generators all in line.
I would imagine that any big server farm would benefit from this kind of setup. Especially when you have people runnnig the lines that are as good as some of the guys in the telo world, they can really make the wiring look like a art in some places.
...this may or may not save a step.
:)
However, it does provide a few significant advantages.
Telcos use DC because it's easy to battery-back. Since all your gear is already running from the DC supply, there's no guesswork about whether your UPS will be able to handle the load. Each piece includes its own converters, so all you have to do is size the battery bank. Since most telcos aim for 8-hour runtimes on battery (long enough to discover and fix a generator problem), overkill is the order of the day.
There's also the point that you can run several small generators, instead of one large one. In an AC world, keeping multiple generators syncrhonized is nearly impossible on a small scale, so you just run one big one. If your setup grows, you rip out the old generator and replace it with a larger one. In DC, since all your generators feed the same battery bank, you can just tack on more capacity without trashing your original investment.
Using multiple generators provides cheaper redundancy too. In an AC setup if you wanted to be protected against a generator failure, you'd need two identical gensets, each large enough to run the whole load. With DC, say you had 5 generators but 4 could power the load. You still have no single point of failure, and you don't have to buy *double* the generating capacity.
Oh, and if a second generator fails, say you're down to 3, you're below the break-even point, but you're still limping along, with the operating generators assisting the batteries, extending your battery runtime long enough that you can probably fix one of the failed gensets. Oh, you found a spare generator at the rental place down the street? Switch a few rectifiers onto it and watch your charge status come back into the green. You just don't have that sort of versatility with AC.
DC is easier to noise-filter than AC. Keeping the high-frequency noise from switching converters off the AC input is something of a black art, and is hard to do effectively. You also have Power Factor (PF) issues when running large numbers of computers (or anything that uses switch-mode power supplies) from AC. Hence, your supplies have to be PF-corrected, which adds bulk and complexity, and reduces efficiency.
A DC-DC converter suffers none of those problems, going from your 48v battery bank down to the 12, 5, and 3.3 levels in your servers. It's easy to filter the switching noise because the input is DC, a big L-C filter works quite well. There's no such thing as power factor on DC, so the converters themselves are simpler and smaller, and run cooler.
One other huge benefit is that 48 volts is "low voltage" according to the NEC, so you can wire it yourself. You'll never have to let pole-climbers into your server room again.
Another advantage is that most DC-input equipment has a telco heritage, and supports dual inputs. Everything in telco has an "a-side" and a "b-side" power supply. It's only relatively recently that high-end datacomm gear has started to support multiple AC power inputs. History and experience are on your side with DC.
Either way, there is still only one conversion: 110AC -> 12/-12/5vDC.
No, there is one conversion 110AC -> 48V DC.
Then, individual pieces of of equipment use DC->DC conversion to get whatever voltages they need.
Old people fall. Young people spring. Rich people summer and winter.
DC power in the data center unquestionably makes sense. Higher density, less heat and if it is on UPS, a lot off efficiency gains both at the computer part and the power distribution systems. Power distribution from the UPS is generally cheaper not having to transform from battery to A/C current. It would even save on air conditioning costs through lower heat on the computer and UPS electronics/transformers. In the long term, such a data center would be much less expensive to operate.
But the only draw back is equipment availability which would cause you to need both power systems in the floor and rack. More cable is always a pain. So if your major supplier provides for this, it is worthwhile.
On virtually every benchmark I've seen, 2 way dual core Opterons (4 cores total, the most energy efficient configuration) totally smoke similar Xeon configurations in performance, while using far less power. WTF?
Is this publication owned by Intel, like Cnet.com.com.com.com?
The argument for AC to DC conversion on each device is that individual power supplies provide isolation e.g. no direct current path from DC ground of one device to another device. This is why Ethernet is transformer coupled. Eliminates ground loops and propagation of damage during major hardware meltdowns, nearby lightning strike etc.
As long as you're feeding a site with AC (this is the only efficient way to transmit it from the Hoover dam to some farm in Iowa), then at some point, there is conversion from AC to DC before it gets to the circuits of any computer system.
Where there *could* be some benefit is where you have larger, more efficient converters very near the point of use. If you figure each power supply inside each box is 50% efficient, but a single big one is 75%, then you reap a net benefit (totally rhetorical - I have no idea how efficient they are).
Now, even if there isn't an efficiency gain, there are two other reasons to do this, which is entirely why telcos do it (not to save power). First, telco equipment has to run when utility power is lost. Every switch is powered off DC that comes straight from BATTERIES. There is no loss in service if AC power is lost, because the batteries are already in use. This explains why you can use your POTS line even during a brief power outage. We have such a box sitting in a machine room in attached building.
Secondly (and perhaps more germane to the concept of DC in non-telco datacenters), telco equipment is often housed in locations that have poor environmental controls. By positioning the not-so-efficient AC-to-DC conversion hardware away from the sensitive electronics, you reduce the heat load in the climate-controled space. This alone makes some sense for the direct DC feed concept.
Charles
There exists no way of exchanging information without making judgments. --Bene Gesserit Axiom
The first step will be in the home office. Have you taken a look at the rat's nest under most desks? Most of it is AC/DC conversion. If the industry could just arrive at a DC power standard, we could start with a single AC/DC "box" under the desk with a standard plug end for all DC peripheraps. Add daisy chaining and wireless USB or Bluetooth, and that nest is largely eliminated.
At this point, we could start to build it into houses and other buildings.
More
Conventional system:
AC -> Building -> UPS -> DC converter -> Battery -> AC converter -> Server Power Supply -> Components
DC-only
AC -> High efficiency DC conversion -> Battery -> Server Components
More
1) Contacts tends to rust on the positive side.
2) Lower voltage means bigger current for the same power. This would require thicker, more expensive cables
3) DC-DC voltage conversion is, somewhat less efficient... Ok, I know switching mode power supplies are efficient, but this leads me to the last point:
4) No insulation between systems. That way, systems get more prone to ground loops...
"There is always an easy solution to every human problem -- neat, plausible, and wrong."
H. L. Mencken
Rectifiers, the devices that turn AC into DC, AKA diodes, are cheap and easy. So there wouldn't be a problem running the system.
PoE compliant switches are actually rather complex and do a great job at providing safe DC electricity and cutting off shorts. As an EE I read the standard expecting it to be a total bodge-job like all else ethernet related, but I was actually impressed and wouldn't hesitate to use it.
Now if we could only get all our gadgets and gizmos to burn less than 14W...
Someone had to do it.
Yes, central DC is attractive from a number of viewpoints: easier redundancy, higher efficiency, integratable into UPS and perhaps most important these days -- removing headload from CPU bays. But you're gonna need big busbars, especially if your boards need 3.3V. I can't see using some sort of non-custom miniPSUs to do12VDC to ATX.
There are definite advantages to DC power... but it can also be *hugely* annoying.
I've worked in DC powered labs.
There isn't really any concept of 'plug' in the DC powered world. Powering up a device usually entails reading it's current draw off the equipment, selecting the correct gauge of wire, cutting the correct length of wire, strip both ends, hook up to your DC distribution on one end and your equipment on the other, select about the right size fuse, plug it in... etc. It's a royal pain. Oh, and make sure you do it correctly, because it's not that hard to electrecute yourself...
Nearly every engineer I've ever worked with whose been exposed to DC powered labs has begged to return to the AC powered world... it's just MUCH easier to work with.
On the flip side though... telco racks rock! Nothing beats hex head rack screws... you can literally drive them in at a 45 degree angle with a power drill and it's OK. It makes going back to the world of crappy philips head data wrack screws that you occasionally have to drill out because the head has stripped very annoying.
Oh, FRANCE telecom. That says it all! Ribbit ribbit!!
Is that DC power is naturally unstable. As loads fluctuate, the conversion and distribution system can change dramatically and result in very unclean power. If you are proposing to ditch AC in the server room and run DC from the UPS hardware directly to the rack, you will need to add in a lot of hardware to guarantee that the servers get exactly the voltage they need. This hardware will probably be less costly and wasteful than the AC systems currently in use, but they will also be more proprietary and (in the short term) more expensive to buy into. This is not the magical solution many envision, but it has a good future since transistor technology is getting a lot better and hence voltage management will be easier and easier as time goes on. The opportunity to move the conversion heat away from the inside of the server allows for better heat management, since you can let a transformer/transistor power system toil away only cooling it from the air duct on the roof and save the crisp AC for the servers.
Actually the 50-60Hz of household AC isn't all that great for power conversion by today's standards. People either rectify it and throw it through a second higher frequency oscillator or use a custom active-switching DSP solution.
Someone had to do it.
Wouldn't you still need a power supply to convert to the various voltages required by a computer, which may change over time. Most DC power setups I've seen run at 48v, which still requires conversion to 12v, 5v, and 3.3v. You can buy a 48v power supply for most servers and other equipment today. With a switched power supply, you'd need larger capacitors or a higher switching frequency in order to smooth out the lower powered DC. It's very unlikely that you would eliminate any heat loss. I would assume that telecommunications equipment uses a 48v setup due to legacy issues and that it was a better idea before switching power supplies became cheap and efficient.
.8 or so, meaning 20% of the billable power is effectively wasted and could be recovered for a slight increase in cost.
The downside with DC is that lower voltages require much thicker wires, and you're at much greater risk for fire. Circuit breakers and other things are also more complicated and expensive since DC tends to weld things together.
An EE I know just built a data center supplying 208v (2 branches of a 3 phase iirc) to all the racks. Almost all existing power supplies can take it, and it saves a bundle in wiring costs. I'm not sure about servers, but most desktop power supplies operate at a power factor of
There is already such a protocol. For devices > 14W a second standard would be needed, but it would just be a tweak to the current standard to change wiring/values to something more suitable for large appliances.
http://www.poweroverethernet.com/
Someone had to do it.
This is similar to an idea I had a while back, which is if I were in charge of the data center for a lean dot com, I would install laptops in the datacenter. They have built in UPS for up to about 2 hours of run time. They are powered directly off of DC, so there is no need for DC to AC conversion coming off of the DataCenters UPS. I suspect that Laptops would generate less heat, and be reasonably power efficient. Lots of cheap laptops on trays might be cost comparable to 1u server boxs at the same density.
This actually saves 2 conversions. One from data center UPS batteries to AC, and then back to DC in the server.
Think Deeply.
A little research goes a long way.
Notice all the references to 48V.
Why such low utilization ?
Any other industry would scrap 80% of that equipment to save costs and power.
And if you would have read the next sentence instead of knee-jerk posting, you would have learned that they are doing exactly that.
Not affiliated with them, just like their half-depth 1U servers quite a bit.
Their DC power stuff is quite cool:
http://rackable.com/products/dcpower.htm
-Chris
I've been saying this for years -- outfits that already have a DC infrastructure are natural candidates for photovoltaic augmentation. Slap a few panels on the roof, the rest of the system is already in place. Why not?
I'm not going to suggest that a telco CO could run entirely from rooftop solar -- far from it! But they'd see a much faster return on investment, compared to residential systems, because they wouldn't have to buy inverters, charge controllers, or any of that crap. (One advantage of an undersized array is that you never have to worry about overcharging, so you don't need a charge controller. Maybe MPPT would help efficiency anyway.)
I doubt many people would be happy with the large copper bus bars that would be needed to distribute low voltage DC at any reasonable power level.
Mea navis aericumbens anguillis abundat
Assume that an AC-to-DC conversion causes a loss of 10% (just to have a number).
If we bring in AC, convert it to DC in one location, and then distribute it as DC to all the computers, we've lost 10%.
If we bring in AC, distribute it to all the computers, and convert to DC at each computer, we lose 10%. The conversions are independent and parallel, and so the loss is not additive. (After all, if we have 10 computers, it doesn't mean we are losing 100% of the power). I can see how we might save money, as we no longer would need a complicated power supply at each computer. Also, we wouldn't have a hot power supply in each computer, and this could reduce cooling costs. But I don't see where the power savings comes from.
The slashdot story intro implies that the advantage of DC is that you
save a conversion step. Well, maybe you do, maybe you don't, but
counting the number of AC-to-DC and DC-to-AC conversions is very
misleading.
Converting 50 or 60 Hertz to DC is much more costly and less efficient
than converting in either direction at a higher frequency. Low
frequency rectification requires large filter capacitors, complex and
expensive inrush current limiting, and active power-factor correction.
By doing that front-end work in one place only, preferably from a
3-phase source, you save power and increase reliability. You probably
still want multiple 50/60Hz to DC rectifier stages, of course, but now
they can be in parallel (for redundancy), rather than each one
downstream of the other where a failure of either one will bring down
the system.
Just because you're distributing DC to the racks, doesn't mean you
don't have to convert it again. It typically gets converted to AC and
back to DC at least once, usually twice before it reaches CPU and
memory chips. That's equally true in data centers that distribute AC
or DC. The fact is, memory and CPU devices want very low DC voltages
and very high currents. To make matters worse, not all parts of the
system want exactly the same DC voltage, you almost always have to
have multiple supply rails. You can't distribute very low voltages,
because it would require wires as thick as your arm and they'd still
be too resistive and inductive, so instead you distribute the DC at,
typically, 48 volts. The subsequent conversion to low DC voltages has
to happen via an intermediate AC, but it's a high frequency AC, so it
can be done much more efficiently using ferrite magnetic components,
active rectification, and often resonant mode filters. This high
frequency AC is confined to the internals of a power supply unit, it
never travels over wires or between boxes, thus reducing typical
high-frequency problems such as RFI.
I haven't mentioned battery-backup (i.e. UPSs). They make the system
more complex, but don't change any of the fundamental concerns. Even
on a DC distribution system, the UPS system requires it's own
additional stages of DC->AC->DC conversion, both while charging
(standby) and while discharging (during AC power failure). This is
because battery charging has to have a precisely controlled current
envelope. And batteries don't discharge at the uniform and
well-regulatted voltage that your DC distribution wants. They need
regulators, and switchmode regulators (typically DC->AC->DC) are the
most efficient choice.
I think I'm going to cry. Someone named Nigel working for the French called me mean names.
I need to find my tissues....
I'm just a caveman programmer. I don't understand your strange, "modern" ways of thinking.
Transmitting DC over long distances doesn't work very well
i um/PACIFIC.htm
I've heard this before, but I haven't heard a terribly good explanation for why.
HVDC Pacific Intertie between Oregon and California:
http://en.wikipedia.org/wiki/Pacific_Intertie
http://www.transmission.bpa.gov/cigresc14/Compend
I worked on a transmission sales automation project at BPA, and I seem to recall some very good explanations for why they had a 2000+ MW, 400kV transmission hop.
Not sure how you figure it saves power since you still use the bulk of the power supply logic to convert the voltage from the 48V supply down to the 12V, 5V and 3.3V that the server uses... And still have to convert the incoming mains power from AC to DC somewhere in the building.
Delivering power at 48V instead of 120V means you're dealing with 150% more amperage... Transmission loss generally follows the amperage. That's why they boost the voltage on the high-power lines. And its the amperage not the voltage that kills you. That's why many folks survive high-voltage lightning strikes.
You'll also pay three times as much for the power supplies since they're not made in the same quantity as normal PC power supplies.
There are only really two major benefits to DC power systems:
1) AC power induces currents in any nearby parallel wires, such as your network cables. DC power doesn't.
2) Battery systems are trivial to implement in a DC power system: simply connect the batteries in parallel with the main supply with some simple cutoff circuitry for if the voltage drops too low.
Moderating "-1, Disagree" is simple censorship. Have the guts to post your opinion.
True. But low voltage (under 50vDC nominal) doesn't require licensed electricians to run it. Clearly the extra buck for thicker copper outweighs the cost of paying an electrician for 8 hours to come extend a power feed. You've obviously never had to deal with licensed electricians.
How so? AC-DC switchmode power supplies start by rectifying the AC into a high DC voltage, and then perform internal DC-DC conversion to produce the output voltages. Even the ones that work straight from the AC are no more efficient once you include power factor correction.
Not true. Remember I said the telco power system is -48vDC with respect to ground? All the logic levels (12, 5, and 3.3v) in the cards are positive just like you're used to. The DC-DC converters are isolating; all they stipulate is that there be less than a 300v total differential between the inputs and the outputs. You're free to reference any part to ground, or leave it floating if your heart desires.
Telco grounding is insane anyway. Most places have #6AWG from each rack to a 1/0 aisle ground cable, and all the aisle grounds meet on a 750KCMil that runs the length of the building, over to the "office principal ground point". Track down a copy of TP76200MP and read up.
Computers should go back to being analog.... then there would not be the extra step of converting to DC. And since the high fidelity folks swear by analog audio systems, just think how much better it would be if a computer used the same.
My karma is not a Chameleon.
Have you ever wondered what is it that makes you so insecure that you can only raise your sense of self-worth by putting others down? It's worth pondering.
From a "Senior Hardware Engineer" of a National telco, it's fine to say people are wrong, but where's your constructive criticism? Where is your backup technical statements saying "this is wrong because...". Instead, you spend 3 paragraphs putting down others and re-enforcing the French elitist stereotype. Then you wonder why people feel so mistreated & revolted they have to vandalize Paris to wake you up.
I was actually impressed by some of the great comments posted by slashdot readers on this article. If you've got some technical information, let's hear it, Eletrical Engineering is something many slashdot sys admins could learn a lot about from people like you, albeit the tolerance level notch needs to be turned up, before you start writting.
adeptus
No trees were killed in the making of this post; however, many trillions of electrons were horribly inconvenienced.
He's in bed with the energy companies, who want to keep their record high profits.
Nuff Said.
It seems to me that it would be ultimately more efficient to just supply AC power to the motherboard and let it create the various voltages that it needs. The same goes for disk drives, etc. In fact, maybe all computing equipment ought to have been standardized on a single "safe" AC voltage, like 12VAC. Then one simple transformer could supply all computing needs without having to figure out how to distribute "clean" DC power. I'm sure it's far too late to make a change like that now, though.
People are cheap. When given the choice between a smart bus that unloads cycles from the CPU, guarantees bandwidth, has more bandwidth, and basically does what a next-generation bus should do, people chose an updated version of the serial port because the latter cost a dollar more to manufacture. I'm talking about FireWire vs USB. The iPod, which practically drove people out to stores to buy FireWire cards and clamor for it on their new computers, now in its fifth generation has dropped FireWire support because... that dollar was just too much.
Cheapskates...
what?? Since when?? That directly condradicts this /. article:
"'AMD currently offers the most attractive dual core option. The Athlon 64 X2 3800+ may cost $87 more than its Intel counterpart, the Pentium D 820, but the AMD chip is a much better performer. It also uses considerably less power.'"
my karma will be here long after I'm gone
Any quality large UPS (generally >10KVa) will be a true on-line double-conversion unit. Even in normal operation, these trickle-charge the battery, and run their load off the battery. There is zero cut-over time on AC fail and on AC restore, and running off batteries provides significant power conditioning.
These systems DO convert AC->DC, then DC->AC, 24/7.
The article on raised flooring was an interesting question, but stupid solutions.
That article talked just like some "Intelligent-Design" moron. Just because HE can't figure out how to properly model raised-floor airflow, it must not be possible to do it at all. Wrong. There are any number of companies that will do this for you.
The solution to raised floor airflow is proper modeling of the equipment, vent tiles, and blowers, and relatively unobstructed floor plenum. The solution is NOT air-cooled equipment on bare floor and overhead cable runs. If cooling is still a problem, then use liquid-cooled racks and equipment. (This is where things seem to be going right now.) While overhead cable runs may work fine for some dinky test lab, "real" equipment requires power cables of a size that would quickly fill most overhead runs.
This article proposing DC power is equally stupid.
An enterprise storage box, fully configured that I looked at requires 13,800 kVA of 208V three-phase power (100A inrush current). My mind can barely fathom the completely unbendable copper "wire" that supplying that much juice at 40-ish volts would require.
Telco's switches have a far lower power density than modern servers, and the DC power was made to correct for different problems.
If this guy's ideal data center is overhead cable runs, ceiling blowers, bare floor, and DC power, I'd run away fast.
SirWired
Easy.
First, DC actually is better for transmitting power over long distances. AC current tends to concentrate in the surface of the conductor, leading to higher current densities and larger ohmic losses.
So, why do we use AC almost everywhere? Transformers. It is relatively easy and efficient to use a transformer to change voltages of AC power. For large electrical lines, the voltage is cranked way up, which means the current is reduced. The less current, the smaller the losses due to resistance in the wires. So power is transmitted at high voltages, so the current and hence losses are low. Then, near the place where power is needed, transformers change the power to lower voltage, higher current. (This is because you can't have house wiring and appliances that won't arc or explode when hit with 13,800 V.)
Converting between high and low voltages with DC power is much more difficult, and requires more complex equipment. (An AC transformer is two pieces of wire wrapped around a chunk of iron.)
Hey Johnson, we are running out of power in our datacenter!
Ok, I'll order more HP Xeon servers. They use more power, cost more, perform worse, and have a limited upgrade path. But Intel sends me cool swag so I use them.
Johnson, your fired! We are going Opteron, and raising our capacity 30%!
ignorance is bliss. googlefiberatx.com
Once the voltage is stepped down for computers with a well-sealed power supply, it's safe. However, you still don't want to go in and mess around with the insides of a computer power supply. Not unless you know someone you don't like who is willing to help :-).
win servers are not wasted. they are needed. a win server at high load, just tanks, and you get calls all day, why is mail slow, why is printing slow, why does it take over a minute to open a file?? yada yada you really need one server per service in an environment of any size (500 workstations), plus redundancy. So at least 2DC's a clutered mail/print/file system... ISA... ePO... cluster manager... back up...the list goes on, and so does the addition of win servers.
So in the end win servers eat a lot of juice... the one unseen cost in TCO.
Sig Hansen?
Does anyone know how Google powers their massive clusters (AC/DC)?
Their way will, of course, be the "one true way" to supply power.
I read
I'm surprised that no one as mentioned that. If things are DC powered, then the DC of the back-up batteries doesn't need to be converted to AC. There's a power savings right there when you REALLY need it.
It's related to the other issue the poster mentioned, that AC can be converted easily and effeciently between low and high voltages. Being able to transmit the power at effecient high voltage and then step it down to a much lower voltage for use without losing much in the conversion is what makes AC more effecient.
I used to work at a large Telecom that had DC power. Every month two techs would go check the power room. It involved all sorts of fiddling. I asked them what that wooden bat was doing hanging at the ready. "That's for if you get stuck. DC won't let you go. And you'll need the other guy to wack you with the bat and save your life." They were probobly generating and storing a lot more DC than most. But it's still something to keep in mind.
Just about everything that has an electric motor runs best on AC. Your Air Conditioning, Washer, Dryer, water pump if you have one, fans, table saw, drill press, and so on.
You can have DC motors but they tend to be a lot less efficient than AC motors not to mention that they tend to have brushes which wear out over time. Now throw in the items that Require high voltage. Your CRT and television if you are using them. It is a lot more efficient to shift voltage with AC than with DC. Then you have the devices that use the AC frequency as a timing signal like TVs and some clocks. The 60hz signal is pretty stable over time so it is often used as a quick and dirty clock reference. That is why TV runs at 60 fields/30 frames a second unless your want to talk about drop frame but let's not go there.
See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
I've thought the same thing. Perhaps DC could be wired in the wall-socket next to the AC plugs, to be backward-compatible. (Presumably people are a bit hesitant to change an electrical socket standard that's been around for many decades.)
The problem, though, would be what voltage to standardize on, and how to step down. Say you distribute 12V. When you have a 9V device, what sort of electronics do you use? Expensive? Power-hungry and hot-running?
Laws do not persuade just because they threaten. --Seneca
Actually as users we could benefit now from this on the user end.
Move power supply functionality into existing UPS's and run a DC harness into your computer.
Also run a harness into your DC LCD monitor.
Vendors would only need to replace their DC->AC converters with 12/5/3.3V output levels.
This should allow for cooler, smaller computer cases.
I think this article might be using the term "engineer" a bit too loosely. I doubt any company would hire an engineer - an actual person with a Professional Engineer's License - to work on these systems. A more appropriate term might be "technician," which usually refers to someone who is trained to repair and work with a single type of technology. Engineers, on the other hand, are usually trained to work with a large variety of technologies and usually work on either (A) Research and development, (B) Manufacturing, or (C) Failure analysis and redesign.
I guess using the term "engineer" sounds better though since it tends to scare the corporate fat-cats away from a technology because of the implied additional cost from hiring an engineer as compared to a technician.
I'm in the elevator business and we run into that a lot.. I like it because it's actually 120vac from each leg to the neutral. In our business it can eliminate control transformers that supply the 120 to operate the control devices. Controller ground = the neutral.
It's nice, too, when you rectify it as you get a relatively low ripple 275 dc without further filtering, which is convenient for running the door operating motors and the drive motor fields if you are using a DC drive motor.
"Do the Right Thing. It will gratify some people and astound the rest." - Mark Twain
Those battery cables are large for two reasons: (1) High current capacity. (2) Low voltage drop.
It's all about ohm's law. Say you have a circuit consisting of 1000' of 12AWG copper wire (that's 500' out and 500' back). (US standard for 20 amp power service, IIRC.) This wire has a resistance of (~1.62 ohms)/(1000 feet). Push 10 amps through that 1000', and you get a voltage drop of (1.62*10)=16.2 volts in the cable alone.
The drop is strictly a function of the current you push through the wire-- regardless of voltage of the supply. So if the mains supply is 120 volts, you'll only see (120-16.2)=~103.8 volts (~13% drop) at the load end of the circuit. Guess what happens if you drop the supply voltage to 12 volts: (1.62*10)>12, so you can't pull 10 amps through that 1000' of wire... and if you try, the volts delivered to the load end drop to zero!
This is one reason that audio buffs like welding-cable-size speaker wires: Speakers are low-voltage high-current loads, and with high currents the cable resistance becomes significant. Put a standard 8 ohm speaker at the end of that circuit above, and you'll use about 2/3 of your amplifier power to heat up the copper.
So it's not just a matter of the current the wire can safely handle without melting: At low voltages you have to be sure the wire is big enough to keep the voltage drop within reason. And that can get expensive: If all you can accept is a 10% (1.2 volts with 12 volt source) drop, then to handle a 10 amp load you'll need a cable with (.12 ohms/1000') or better for that 500' run. 1AWG is ~.126 ohms/1000' and is also ~.29"/~7.3mm in diameter... lots of copper!
No, electric chairs use AC because Edison wanted everyone on DC (Guess where he had invested a lot of money). To create fear in people's mind he invented the electric chair, which he ran on AC in demonstrations. Because it is impossible to get DC to high voltages (with the technology of his day, now we get DC just as high), his chair could not run on DC, which he convinced people made DC safer.
If you are strapped to the chair it doesn't matter that there are zero crossings with AC where you can pull away - the straps keep you in place. With DC there are none and your muscles don't obey you.
There are several approaches to power distribution. One is "telco type" -48VDC distribution. This is most appropriate when the configuration doesn't change much. Wiring usually involves big cables and screw lugs. Plugs aren't standardized. More importantly, there's no set of simple rules, like the UL/NEMA/NEC standards that govern plugs, outlets, wiring, and circuit breakers, that make 120V power distribution safe without having to measure everything.
In the 120VAC world, everything has been designed so that end users don't have to worry much about overloading the wiring. If they do, a circuit breaker will trip. An ordinary power plug, a "5-15P", can handle 15A, so if you have an outlet strip, there is a breaker to protect the plug and cord from overload, should the total load on the power strip exceed 15A. A 20A power strip must have a "L5-20P" plug, the big twist-lock type. As soon as you get away from 120VAC, you lose that designed-in idiot-proofing. (Europe is still struggling in this area, with too many different connectors, so you don't get the same level of idiot-proofing in the 220VAC part of the world.) So once you leave 120VAC, you're going to need power engineering skills. (Clamp-around ammeters are very useful, and yes, you can get them for DC.)
There's also 400Hz AC distribution, which allows for smaller transformers and filter caps in power supplies. 400Hz rackmount servers are available. Aircraft, military, and some mainframe systems use 400Hz. It's not a big win in this era of switching power supplies.
There's 3-phase power distribution. Here's a 3-phase outlet strip. More to the point, there's an efficiency gain in running a UPS from 3-phase power, and big UPSs are usually 3-phase, at least on the input side. Arguably, power should be 3-phase down to the point where it's rectified to DC, because 3-phase rectifiers need far less filtering, but nobody does this for small loads.
American Power Conversion has been pushing the idea of integrating power conversion, cable management, and cooling into standard racks. Classically, those are the big problems in big computer systems. Seymour Cray used to say that the big problems were "the thickness of the (wiring) mat" and "getting rid of the heat". By that standard, APC is now as much of a computer manufacturer as, say, Dell; neither makes motherboards or ICs, they just package gear from others. Which is a wierd thought.
All of this power is going to be converted again, at least once, and probably twice, before it hits the semiconductors. That's the job of point-of-load DC to DC converters, usually ICs on the board that do the final conversion. Typically, when you get to the computer, there's a conversion from the line voltage (120-240VAC, 48VDC, etc) to internal distribution voltages of 5-12VDC, then another conversion and regulation just before each device, usually downward to something like 3.3VDC. This keeps transient load changes from one device from affecting others. There may be on-chip regulation, too. The losses at those last stages of conversion are usually the biggest ones in the whole chain.
Actually DC is better for very long distances.
There is a limit to how much voltage you can use in a wire. Go beyond this and you end up with problems related to the air around your wire. Normally we measure AC voltage RMS (I'm going to do some handwaving over some math and call this average), but this maximum voltage in the wire much be measured at the peak. So 120V RMS will have peaks that are about 170V. DC allows us to stick at one voltage, which can be closer to the peak
AC however is easier to work with. With a simple transformer you can efficiently change voltages. This is much harder in DC (efficiency counts). So where there is a need to go long distances they use DC, then transform it to AC in big complex plants, and run it all over your city at whatever voltage is handy, switching as needed.
I disagree. The contacts on my car to get icky. A chrome lighter plug is good at resisting this, but in my 10 year old car there is a noticeable change. Most auto contacts are covered in rubber which prevents air from getting in and solves the problem. It is an issue though.
It doesn't help that cars are normally exposed to more water than a house outlet.
Solar cells and fuelcells produce DC power. And their architecture is most effective distributed around the grid, for local generation/consumption. We would have a lot more efficient grid if we used DC like we use TCP, and AC like we use IP, for a "Power Internet" (iPowerNet?). Where only internetwork transmission is AC, it's all redundant, and nodes are "prosumers" of power.
--
make install -not war
Our Colocation hosting fascility offered us DC power if we wanted. They said, we can have as many servers drawing DC power as we want (while limiting us with AC power). They said it is up to us but they would prefer if we used DC.
However, our hosting fascility is also a Telco. They operate all their phone infrastructure with DC power. Therefore, they have the knowledge and a very reliable system for that.
The problem was that we had very hard time finding DC PSUs for the devices we wanted to buy. When looking at blades from the top manufacturers, they all had AC only solutions. I thought it was a shame.
Another issue would have been, what if we wanted to move to another fascility? They might not have DC. We would be either limited to DC only fascilities (which is very limited) OR buying AC power supplies for all our machines (waste of money).
I think DC is a great idea. Hosting fascilities should all move to DC asap. It makes sense. But they should all do it, and they should be backed up by hardware manufacturers.
"From the moment I could talk, I was ordered to listen" - Cat Stevens
What I'd like to see is rack-mount systems come with the power supply split off as a separate unit.
This would let me draw power from any DC power supply that was up to the task. If I wanted to, I could put a large PS at the top of my rack or another rack altogether and feed the entire rack from it. If done redundantly your entire rack can be protected against power issues.
I haven't researched this but I'd be shocked if such a thing didn't already exist.
Unlike the system outlined in the article, this won't necessarily save any power, but it can help with heat distribution and can help organize rack space better.
Knowledge is how to play a game, intelligence is how to win, wisdom is knowing what game to play.
There are few different issues:
1. The ohmic losses are equal to R(I^2) so you want the current to be as low as possible. This is why transmission lines are high voltage.
2. I am a little fuzzy on the specifics but the radiative losses go up with higher voltage when using AC. I suspect this could be modeled based on the impedance. At some point, the voltage is high enough that the cost of doing AC to DC and DC to AC conversion is less then the radiative losses. DC transmission lines do not suffer radiative losses.
3. DC to AC and AC to DC conversion is expensive in terms of complexity, cost, and efficiency compared to using transformers for step up and step down.
The result is that the transmission line between Sylmar and The Dalles is both DC and runs at a much higher voltage then an AC transmission line would.
An enterprise storage box, fully configured that I looked at requires 13,800 kVA of 208V three-phase power (100A inrush current). My mind can barely fathom the completely unbendable copper "wire" that supplying that much juice at 40-ish volts would require.
I think you meant 13.8kva, not 13.8 megavolt-amps. 13.8kva at 48VDC is only 288 amps. This is far less than what most telco systems run these days. The conductor size needed to handle this load is no larger than that feeding most newer homes, and is certainly flexible enough for installation.
I am a geek attorney, but not your geek attorney unless you've already retained me. This is not legal advice.
A very righteous post from you. However, you seem to be assuming that that guy wasn't just some troll kid... Which isn't necessarily true, this being slashdot and all, you know? :)
The AACS key is NOT 0xF606EEFD628B1CA427BEA93A9CA9773F
Smaller transformers are the main benefit of higher-frequency AC, you're right. But the 400Hz power system is found plenty of places besides airplanes. I'm not millitary, but I've heard that most shipboard power is 400Hz, and they also do something weird with phases and ground, but I don't recall the specifics.
I understand that old mainframes used 400Hz because the power converters were smaller and made less heat. The easiest way to make 400Hz power at the time was through "rotary converters", a motor-generator pair that takes in whatever power is available locally (50 or 60? One lump or two?) and spits out 400Hz.
I think you meant 13.8kva, not 13.8 megavolt-amps.
:-)
Whoops... yeah, not 13.8 mVA. (imagine a Beowulf cluster of those!
13.8kva at 48VDC is only 288 amps.
What about the 100A of inrush @ 208 three-phase? What would that take with DC? I never did take a power engineering class in college.
The conductor size needed to handle this load is no larger than that feeding most newer homes, and is certainly flexible enough for installation.
Yeah, but that is the current load for a single system. You put in half-a-dozen of these things, and you are talking serious quantities of copper. A quick check of a googled wire chart shows that 300A takes a conductor just shy of half an inch thick. Two of those together x 6 systems (by no means an unusal setup) starts to give me the heebie-jeebies, cable-routing wise. (Do we need a nastily thick conductor for earth ground also?)
SirWired
Hi everyone,
I am a datcenter manager that has had the opportunity to not only run but also build a datacenter from pretty much scratch. In my experience I have found that both DC and AC powered equipment both have their places in the environment. Neither system is perfect so by running hybrid you can get the most flexibility.
We recently moved our datacenter form a 10K sq ft facility down to a 1700 ft facility by doing a technology refresh and changing many of our key infrastructure methods. In the new facility I currently have 315 HP blade servers plus another 10-15 traditional rack type servers running. I have the capacity to add up to another 144 blades (assuming they are 1U) before I run out of floor and HVAC capacity. The power delivery method is hybrid. I run DC for the blades which are fed by Emerson Energy's Candeo XL rectifier stacks (originally designed for telco) and AC for everything else. To eliminate a lot of the under floor clutter I use a trough system instead of conduit for the various AC circuits. HVAC is provided by 4 Liebert 22TON units which keep my room at a comfy under floor temp of 66 degrees.
Adequate airflow is critical so we spent a lot of time planning tile placement. The key for proper cooling in this scenario was a high volume of airflow pushing the cooling to about 5.5ft up from the raised floor. This way my cooling isn't being sucked up by just the bottom half of the rack. Low voltage cabling is overhead.
We chose to power the blades DC for two reasons. First was the limited space I had for installing breaker boxes on the walls. The number of AC circuits I had was limited so I pulled fat feeds directly to the Candeo systems. A full rack of HP p-class blades would require 4 x 3phase 208 circuits per rack. My initial installation of blades would have consumed 144 of my 168 circuits leaving next to nothing to power my SAN/Network/Tape Library/etc equipment. The other reason was power supply efficiency. In the conversion of power from AC to DC the efficiency of the power supply must be taken into consideration. It's not just the number of conversions you do but the loss at each. Typical power supplies in servers run about 80% efficient while my Candeo as they are setup gets about 90%. For me this ultimately meant less heat and more available cooling, therefore I could bring in more servers under the existing HVAC.
I prefer a best of class mentality. IMHO there is no best universal solution. For those of you that use traditional rack mounts servers like Dell you can purchase these units with a DC option. I am not sure if HP offers a similar option but they might.
Len
p.s. Someone also made the comment about DC not generating noise in network cabling while AC does. This is not a totally true statement. Anytime you run a current through a conductor you will generate a magnetic field. Put this in parallel to another conductor and you will further induce a mag field (this is why any power runs that have to intersect low-voltage cabling should only intersect at 90 degree angles to avoid inductance). The big difference is the way DC cabling runs. In most DC circuits the feed and return lines run together so the proximity of the out of phase magnetic fields will cancel each other out. Don't believe me? I had this problem when we intially wired these Candeo systems up. The small feeds to the racks and the big mains that connected to the common buss bar were about a foot apart. Because the fields weren't cancelling, we were getting enough noise on the lines that it looked like there was AC leaking through the circuits (6volts p-p in some cases). By simply wire tieing the lines together, the proximity cancelled the fields out and everything was peachy.
what it means...
A "switching" power supply refers to its internal mode of operation, as opposed to a "linear" power supply. Switching converters are more efficient, but more noisy, than linear regulators.
A universal-input power supply, to which you apparently refer above, is an entirely different animal.
I was referring to the central office grounding procedures, which are partly to deal with conducted strikes coming in from outside, but mostly to deal with RFI, leakage current, and safety ground should some power converter fail horrifically.
Considering that many offices don't even have copper phone lines leaving the building anymore, and the grounding practices remain in effect, I think it's mostly for the latter. I wish Nortel weren't so tight about their installation manuals, you could spend a few days reading their switch power and grounding practice. The SBC documents I linked to should provide plenty of detail anyway.
Point is, it has little to do with pole-strung wire.
No, you convert in one place only.
Bad troll! No cookie!
Just because *you* can't figure out how to run overhead power cable, doesn't mean it's a bad idea. If you're not running your power cables on their own rack, separate from signal cables anyway, you shouldn't even be in this discussion.
That being said, I'm very interested in seeing any datacomm setup that fills a 24-inch power cable rack. You're in the vicinity of a thousand amps there, without stacking things very deep.
You're also obviously unfamiliar with "superflex" cable. It's similar to welding cable, having a rope-lay configuration of hundreds or thousands of strands. Plenty of telco equipment is powered over four runs of 1/0 or 4/0 superflex. (That's a supply and return for A-side, and a supply and return for B-side. Welcome to the right way to do things.)
Again, please don't flame the entire article just because your experience is limited.
Gas everything, I do laundry rarely using an efficient front-loader, my fridge barely gets opened.
I have between 5 and 10 computers running usually, plus an unholy number of other electronic devices. Heck, I've thought about combining my wall-warts with a homemade DC distribution rig (maybe combined with battery backup and solar assist) just because I have 30 or more wall warts plugged in usually. When I need an outlet I struggle through the nest until I find an unused wall transformer
Your point is true for the typical American house, but we're talking on Slashdot.
Man, you really need that seminar!
Ouch. Lets take the context of the article's discussion (which you correctly conclude as 'dribble') into account before we overreact. My statement was based on the conclusion many people made regarding the article, and that is that if they could make one magical huge power supply at the head end, they would need no subsequent supplies at the equipment level. This is false, and the reason is that DC power would fluctuate wildly if converted and transmitted the distances typically in play in a data center (making it unclean or perhaps unstable?). Within a server, tolerances are a volt or less, and you would never meet this without power conversion systems at least at the rack level, if not at every server. But what do I know, I'm not an engineer for a major power equipment manufacturer or anything (or am I).
To paraphrase Rich Teer - who in their right mind puts a frame buffer on a server?
Stick your hand in the power supply and tell me if it's "not high voltage". Of course, the internal components are not high voltage, but you can't have a city-wide delivery system based on 12/5/3.3v, let alone a datacenter.
Actually, power below and signal above is not a terrible idea, I will admit. I have never seen this deployment in the field as most of the shops I visit are ex-mainframe shops (or current mainframe shops), and none of them use overhead cabling (at least for the Computing part of IT. Telco is comething different.) Now-obsolete Bus and Tag (mainframe) cables make hi-amp copper wire look compact, hence the lack of overhead cable runs.
The "average" shop I visit uses power run through conduits bolted to the concrete floor, and then data cables are bundled together and run point-to-point on top of that. Yeah, I know, makes power upgrades real interesting with live fiber cabling nearby.
You're right, I am unfamiliar with superflex cable. My bad. And yes, I know that all real equipment has an A and B side.
I must mention that a mere thousand amps isn't very much. If a single rack requires two L6-30 connections of 208V AC, your 1000A 24" power rack starts to run dry awful quick, especially if you start using lower-voltage DC.
Beyond that, overhead power runs also require a minor re-design of some equipment, most of which expect the power to come in from the bottom.
SirWired
I hate those big power bricks that when plugged in, wont allow a second small power socket to fit. Its too damn
fat.
Cant they either make em thin, or use a wired plug - like laptops.
But what we need , are DC outs at the back of PCs, so I can plug the wifi, or adsl modem and 8port switch into the back
of the PC and use the PCs power.
Liberty freedom are no1, not dicks in suits.
A datacenter is not 1 computer or even 1 rack. It's BIG--it takes MEGAwatts of power. You can't supply power for a datacenter on 3.3V, or even 52v. You need high-voltage to power thousands of cards. The high voltage of DC is dangerous to control and hard to regulate.
You'll lose a finger.
Not funny, it's happened.
So irrespective of what she who must be obeyed has to say on the matter rings and watches must go in pockets before working on high current low voltage systems.
Most of the "new breed" CLECs run AC power and air conditioning below, DC power and signal above. So the data folk get what they're used to, the telco folk get what they're used to. If you'll pardon the inconvenience of bolting racks to a raised floor. And you know the last idiot didn't file the burrs off the ends of his Unistrut.
I know of one MCI/WCOM facility that ran DC power overhead, signal and air conditioning below. (No AC in the switchroom.) Talk about a mess. The DSX aisle had so much cable piled up under it, the airflow to the rest of the room was heavily constricted. This was with an 18" floor height, too. The overhead power looked great, though, with its alternating paired colors. You could tell it was a WCOM facility though, -48 was red, not blue. (Old MCI installations do it right: Blue helps snap you out of the "ground is negative" convention that so many of us get into.)
All the Bell COs are solid floor, so everything's overhead. Depending on the vintage, you might be dealing with 14'6", 14', 9'6", 7'6", or 7' frames. As circuit density has gone up, it no longer makes sense to fill a 14' frame with equipment. Hell, even a 7' frame has so much cable entering it now, you have to put spacers between the frames or just not fill them completely.
Doing everything overhead works fine, except you have to note, AC is only in a CO for convenience receptacles. There are only a few circuits, for lights, tools, and the occasional printer. If you had to put up enough conduit or wiring gutter to support servers, it'd get ugly in a hurry.
As much as I despise working in raised-floor environments, I have to admit, the hybrid over/under layout that the young CLECs use is pretty slick. It accomodates gear from both telco and data mindsets.
I just have one thing to urge you: If you're putting signal cable under-floor, put it in writing that when someone removes equipment, they must remove associated cabling, using appropriate practices to avoid damaging the other cables. Otherwise after a few generations, your floor's full of old cable, packed too tight to safely remove any of it. That can happen with overhead racks, too. It just takes a lot more neglect and a lot more years. (Repeat after me: Cable mining is not fun.)
The answer is well-covered in the wikipedia: http://en.wikipedia.org/wiki/High_voltage_direct_c urrent
Slashdot gets worse every day... Pipedot: News for nerds, without the corporate slant
Regardless of whether AC or DC is used in the distribution, voltage conversions are going to be needed in order for the distribution to be practical.
To use telco as an example, they use 48VDC as their distribution. This is a convenient voltage for a few reasons: it is high enough to keep line loss and cable size within reason, and it is high enough to power most equipment without trouble. However, most devices do not operate on 48VDC directly; they tend to want 12 or 5 or 3.3 or [insert CPU voltage du jour], or quite often all of these at once. It is impractical to distribute these voltages over more than a few tens of feet, as you would need really big wires to avoid excessive line loss due to the higher current draw (ever see what happens to badly installed low-voltage outdoor lighting?). Plus it is impractial to distribute lots of different voltages. So the end devices must have their own power supplies to downconvert the DC voltages into whatever is needed for that device.
So we're not eliminating the PSU at each device. But the good news is that these PSUs can be considerably simpler and probably more efficient.
Some people are also confused as to why power entering the datacenter might be converted to DC and back to AC before it makes it to the PSU on any equipment. Beyond the need to charge the large batteries that support the UPS (which requires DC), some installations use what is often called "online UPS". This means that all power is converted to DC, then reinstated as AC 100% of the time. It is like you are always running off of the UPS. You enjoy a higher level of control over the power quality through this arrangement. Unfortunately, these systems cost a great deal of power efficiency.
A fully DC infrastructure would give you the benefit of an online UPS without the power costs. In a DC infrastructure, you have one large bank of DC power supplies that connect directly to the batteries and to the equipment. While the utility power is present, the equipment is being powered from the DC power supply and the batteries are charged. When the utility power goes out, the batteries are able to handle the load of the equipment. No switch-over equipment is needed; the magic of electricity makes this a rather passive system. This is what the telcos have been doing for decades. The only significant difference between a DC infrastructure and an online AC-based UPS is that the online UPS needs a large bank of power inverters to recreate the AC, which chews up more power.
With an "offline UPS" (like the one under your desk), there is always the chance that it won't switch over; the online system eliminates this point of failure altogether. A DC system eliminates both the switch-over equipment and the inverters -- a win-win!
PC and server manufacturers need to get on the ball and develop DC-DC power supplies for their equipment. It really isn't hard at all. I built one for an old desktop PC that I installed in the car to play MP3s. I just took an old dead AT power supply, gutted it, and replaced it with (albeit inefficient) linear power regulators for +12 and +5, and one low-current DC-DC converter (5 to 12) to give me the electrical isolation needed to create -12. I did this all with parts I had lying about. IBM/Dell/HP/et al. could do this in their sleep, and do it with much more efficient means.
Make sure your sales reps know that you want DC! If enough of us bring it up, they'll build it.
Yes, I know and understand what it is. Including their latent properties.
One thing I've not seen mentioned yet, is how DC based distribution protects one computer from another. If a motherboard somehow fries, shorts out, or whatever, and sends back interference or other nastiness back up the DC power line, could it not affect other computers in the center?
The ACDC conversion in a power supply provides a degree of isolation to reduce this problem.
Comments?
Love many, trust a few, do harm to none.
to handle peak loads. If your e-commerce server is dumping over 80% of the customers when everybody shows up to buy, that means one is losing 80% of potential sales. A IT manager who gives the excuse "But we were trying to save on power and capital investment" is going to have his ass fired when this happens. That's a major rationale behing "grid computing", i.e. rent one's CPU cycles when one really needs them.
Tech Public Policy stuff
What about the 100A of inrush @ 208 three-phase? What would that take with DC? I never did take a power engineering class in college.
The inrush results from charging filter capacitors in the server's switching PSU. With a DC source, such filtering isn't needed (okay, we'll still need some filter, but maybe only 10% as much). That's a major advantage of DC -- all that filtering occurs at the main converter that supplies the DC system, not at each device. There will still be some inrush, but only a fraction of the 750A predicted.
Yeah, but that is the current load for a single system. You put in half-a-dozen of these things, and you are talking serious quantities of copper. A quick check of a googled wire chart shows that 300A takes a conductor just shy of half an inch thick. Two of those together x 6 systems (by no means an unusal setup) starts to give me the heebie-jeebies, cable-routing wise. (Do we need a nastily thick conductor for earth ground also?)
True, the total load of the system will be large -- but no larger than for modern telecom systems. I'm not saying the wires are small, only that these problems have already been solved. Telecom systems run many thousands of amps at 48VDC using off-the-shelf solutions easily adaptable to a data center installation.
I am a geek attorney, but not your geek attorney unless you've already retained me. This is not legal advice.
According to a couple of books I have read, at one point Tesla was on westinghouse's payroll.
He also received free power for a while, until westinghouse got tired of Tesla blowing up his generators and cut him off.
The books could have been wrong, ( admittedly there is a lot of 'legend' out there on him and few hard facts ) but thats what I'm basing my statement off of.
---- Booth was a patriot ----