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?
I thinks it's a really good idea.
If there were several pins, many different voltages would be possible, and a device could even use more than one voltage from one plug (eg, it could draw 2V for a relay, 4V for a power indicator, and the standard AC for the actual thing it's powering.)
By not having to have transformers and big resistors inside all the household devices, there would be huge savings in power, things wouldn't get so hot, wouldn't need such big heatsinks, there would be far less electromagnetic radiation around the place (which is probably responsible for a lot of people getting sick etc), and it's safer too (devices that only need a small DC power source won't electrocute you when you drop them in some water.)
Nope, you save two conversions.
Without DC distribution, you have AC->DC->AC in the central UPS, and then AC->DC in each computer's power supply.
With DC distribution, you have AC->DC in the central UPS, and no conversion in the computers.
You get down from 3 conversions to 1.
"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.
I think the best way to do this is to use a smart bus, reminiscent of USB. You plug in the device, and it indicates "I'd like 5V DC", or whatever, and the other side provides the appropriate voltage. A powerbar would have to say "I'd like 120 VAC" or "I'd like raw AC power", then be smart enough to switch that to the desired voltage for each device.
You could only have one plug on each wire from the smart hub.
Costs would be higher due to all the electronics involved, but they'd come down with mass production. It should eventually be cheaper than running all those little black transformers for every device; they draw a certain amount of power even when the device is off.
The DC step in the middle is so you can chop it into a high frequency square wave. You save power and space because you can vary the duty cycle of the square wave for regulation, and the high frequency allows for smaller components.
I've had enough abrasive sigs. Kittens are cute and fuzzy.
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
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
ever seen someone drop a screwdriver between 5V and 0V buses on a 300A distribution system!?
;-)
No problem at all. Turns immediately to metal vapour and will likely not even interrupt server operation.
And with todays computers that would more likely be 12V/3000A
Of course the human being standinge besides this events may suffer some serious damage....
Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
But westinghouse did believe in it .. who then gave Tesla a job.
But later on he screwed Tesla anyway.
Though admittedly, Nikola was not much of a businessman, which is why while he was perhaps the most brilliant scientist to ever exist on this planet, he died virtually pennyless.
---- Booth was a patriot ----
there would be far less electromagnetic radiation around the place (which is probably responsible for a lot of people getting sick etc)
Since when was non-ionizing EM radiation dangerous? Besides, even if it was, I highly doubt that your computer PSU or little brick power supply for your cordless phone are wasting that much energy as EM. When you consider that most little brick power supplies are running most of the waste is lost to heat and not EM I highly doubt it amounts to much. Even your typical PC PSU runs less then 100-150 watts unless you are a dual processor 15 HD fiend.
I'd be willing to bet that you get a larger dose of EM from your box fan, AC motor or refrigerator compressor.
I want peace on earth and goodwill toward man.
We are the United States Government! We don't do that sort of thing.
I guess I find the amazing part is that with most modern switching power supplies, you can plug them into dc and have them just work... no special kit at all. I think it needs about 98VDC for a standard computer power supply.
In a datacenter, it's much easier to provide DC. Most are set up against a large battery backup (to hold over till the generator kicks in.) So, You just simply don't run inverters. The AC->DC converters handle the rest (which they normally do to charge the batteries.) Plus, a AC-DC converter is "stupid simple" in Electrical engineering terms.
I only see two drawbacks:
1.) larger wires to handle the DC current without significant heating/loss (losses are in the form of heat in a wire)
2.) Making sure your routers/switches are set up for DC. Many switches and routers DO NOT come with switching power supplies.... and need a different power supply to run on dc... but compnies like HP and Cisco have DC power supplies for their kit.
I don't really think a data center needs to hire a "Power Engineer" I mean really, did you hire a new person when you switched from Parell-IDE to Serial-IDE??? (oh wait... I shouldn't ask questions I don't want answers to.)
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.
You sir are incorrect. AC is dangerous because the alternating current causes you to grab tight on an object unable to let go. DC is far less dangerous, because it literally causes an explosion that knocks you free of contact with the source. We just went through this in lighting class at my school, and were discussing this as part of why major studios use all DC lighting, its far safer than AC lighting if there is a short.
Yeah, Electric Chairs used AC power because it's LESS DANGEROUS, right?
Slashdot gets worse every day... Pipedot: News for nerds, without the corporate slant
You should look into the Anderson Powerpole plugs. They're made for DC, and can be assembled in a few ways to prevent accidental mixing of voltages. They've become a standard in amateur radio and emergency communications circles, for moving 12v around without the problems of large busbars or cigarette lighter sockets.
The world could use a set of powerpole orientation and color standards. Hmm.
I wonder how much energy a household could save if we were all using DC outlets instead of AC.
Well, none. Assuming you still have a washing machine, dryer, refrigerator, (possibly) electric heating, (possibly) electric oven, hair dryer, (possibly) electric water heater, and air conditioning, which in total consumes almost all of the power that is used in your house, and all of which is more efficient in AC. DC would benefit your electronics, which make up a much smaller portion of the total electric consumption. Some things can swing either way, but as a whole, your energy efficiency is much higher with AC than with DC. AC is much better at running motors than DC is. The real issue that AC to DC conversion has with respect to efficiency is that it produces harmonics on the power lines, which are undesirable and require some damping. The more AC to DC we use, the worse the harmonics are, and the more expensive it is to dampen them. But that's a really really small issue, in comparison to switching everything to DC.
I submitted this story last night, and it didn't get posted.
A typical electrician knows almost nothing about DC power.
This may be changing with the increase of large UPS installations but you are better off with someone that specializes in DC.
Joe and Son's Electrical may have no problem installing a new breaker panel in your basement but don't let them try to run -48v to your server room.
I speak from experience.
don't be so sure... Living bodies are incredibly complex electrical devices. I think it arrogant to assume artificial electric fields cannot have an effect on their proper operation.
(emphasis added) From an interview of Robert O. Becker, M. D., who was a pioneer in the study of natural electrical currents in the human body.
Learn the rules so you know how to break them properly.
www.teslabox.com
Well it looks like you've only started to answer the important question: What do I have to do to vaporize the screwdriver? Here's my stab at the problem:
/V
Boiling point of steel: 2500 C
Specific heat of steel: 0.11 Kcal / Kg C
Weight of metal in screwdriver: 0.1 Kg
Needed change in temperature: boiling point - room temp = 2500 -25 = 2475 C
Q = c*m *( delta T)
where Q is heat added, c is specific heat and delta T is change in temperature.
So, Q = 0.11Kcal/Kg/C * 0.1 Kg * 2475 C
and a bit of arithmetic leads us to:
Energy needed to bring to a rolling boil, Q= 27.225 Kcal
A little bird (OK an online unit converter) told me that this is about 114,000 wattseconds
Now we're getting somewhere...
Lets assume a contact time of 0.1 s. (This is clearly a maximum, since we don't want the disappointment of the screwdriver melting and losing contact) Our power is now 1140 kilowatts- or approximately 1000 toasters worth, or, perhaps 10,000 computers which convinces me that the errors that you fellow Slashdotters have found in my math above are neatly canceling out.
To conclude,
Power = V * I (where I is current)
so,
I = Power
Since we're using 5V
I = Power / 5 V
I = 1140 kilowatts / 5 V
I = 228000 Amps
So, the proposed 300 Amp power supply is off by a factor of 750 or so. We'll all have to head to one of Google's data centers to try this one out. Anyone out there able to arrange a guest tour?
-Jon
PS I know that there is some heat of vaporization to consider, as well, but that will be left as an exercise for the Slashdotter.
You almost seem to be arguing for power distribution to be higher frequency AC, so that the DC conversion componenents can be small, cheap and localized to the point of use. If the HF AC is well chosen for voltage and frequency then some of the intermediate conversion steps can be skipped as it will be "ready" for step down and rectification.
50/60Hz originally was chosen because the low frequencies were easier to generate with the mechanical equipment available 100 years ago but also to avoid a lot of inductive loss into every bit of ferrous metal along the right of way; of course ridiculously high frequencies would actually radiate power away.
I think this could actually work if the power cables were balanced and jacketed at all well, and the voltage was high enough (24v) to carry w/o huge cables. The frequency should probably be whatever switching PS's use these days, ~30Khz?
Also I wonder if there is an increased electrocution risk from HF vs 60Hz.
Especially if you can swing a 12-pulse rectifier - which gives much smoother DC and less harmonics on the AC side. This is becoming less of an issue with PFC SMPS's.
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.
The big reason for multiple conversions is that the latest and greatest chips use ~100W of power at a bit over 1V - so you 100A from the power supply - there is no way that you can transmit that much power at that low a voltage for more than a few millimeters with standard PC board traces. The solution is to use a buck converter right next to the CPU - and these converters can be very efficient (they have to be to stay cheap).
What a lot of people are missing the baot on wrt DC/DC vs AC/DC conversion is that it is easier to get high efficiency with low voltage parts. The Rdson of MOSFET's rises rapidly when the Vds rating goes above 60V, silicon Schottky diodes are only good up to 200V (although SiC Schottkys are starting to show up). The fact that the input voltage is more or less constant means a lot less filtering and no need for PFC.