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

 



Forgot your password?
typodupeerror
×
Hardware

Low Power Servers & Desktops? 28

dhart asks: "Does anyone make low power servers or desktop computers? Couldn't this be accomplished with commodity parts designed for portable PCs. Energy efficiency is environmentally friendly, with the added benefit of smaller and cheaper UPS and AC units in a server setting. If the demand for these units increased, it could lower the cost of energy efficient parts and, by association, the portable computers that drive the technology."
This discussion has been archived. No new comments can be posted.

Low Power Servers & Desktops?

Comments Filter:
  • I suspect this won't happen. The reason is that there is (currently) insufficient motivation for people to consume less power than they need to.

    Power seems to be cheap, a few pence/cents/whatever to keep lights running, run a computer, etc does not concern most (enough) people.

    Laptops are different, since high power usage has a clearly perceived negative influence (shorter battery life).

    Don't know what the answer is. Until we get lots of lovely clean (well, mostly) fusion power we have a problem.

    But generally relying on people's good nature isn't enough motivation to pay 0.5% more for a 'greener' PC.

    Not wishing to bait flames, but I think power is underpriced. Currently, power producers get to consume shared resources for little or no cost - hence the power consumers do not have the cost passed on to them.

    Perhaps one day the idea that consumption of 'fresh air' (via pollution) (and any/all other items which get used up) is a privilege which should be paid for will seem as natural as the idea that you might need to pay for building/living on land. (Perhaps a poor analogy, since I suspect territorial humans have always had a sense of the value of land...ah well).

    Basically, its the 'tragedy of the commons'. There is a shared resource which is being (over-)consumed because it isn't owned and charged out by someone and because those consuming it aren't being lent on by a bunch of angry villagers with pitchforks (I really need to work on these analogies).

    If you throw into the mix that pollution in one country can cause acid rain in another (just as pollution in a river in one country can turn up in another downstream) and you have a hell of a mess which even per-country government intervention isn't going to dent.

    I hope the free market types agree that something needs to be priced fairly (would it be a sensible market which prices a can of beans according to the raw cost of tin and neglected the cost of the beans, manafacture, marketing, etc) and that the green types see that simply imposing national taxes on certain types of consumption doesn't solve the problem.

    So...what should be done?
  • Well,
    in a way, this problem will solve itself, over time.
    The fossile energy-resources will be used up very quickly. Or at least those, that are easy to exploit.
    Then, energy will get way more expensive. Unfortunately, this might crash the whole system we live in, so there might be some adverse effects...

    This is all very obvious. Just consider all those devices that will end-up always-on when the ip-space is large enough.
    Or think what would happen if China had the same car-density as US or Europe (>1 car on two people).

    With the current system, I see no way out, unless energy gets way more expensive.
    (and here in Germany, it is expensive already, thanks to "Green" government. But if you don't offer an alternative, people will really just go mad - witness the strikes/blockades of UK/French petrol-stations)

    cheers,
    Rainer

  • rainer_d:
    Well, in a way, this problem will solve itself, over time. The fossile energy-resources will be used up very quickly.

    At present usage rates, the U.S. has 800 years of coal left [stanford.edu]. China's coal supply dwarfs the U.S.'s. Oil can be produced from resources not yet tapped such as oil fields with a higher cost of extraction, oil shale, tar sands, and the aforementioned 800 years of coal -- so dwindling conventional reserves are a non-issue.

    and here in Germany, it is expensive already, thanks to "Green" government.
    Thanks to Green government, Germany's green nuclear power program is being decomissioned.
  • by Anonymous Coward
    I agree totally on the need to find measures that will limit the the growth in use of fossil fuels in electric power stations. On the other hand, taxing the electricity per se may not be the best solution, at least not everywhere and all the time. Reason for this is that, in certain parts of the world (i.e. here in Norway) household electricity is based (some 98% i believe) on hydropower which is not at all polluting. Likewise, in other parts of Europe (Denmark, Netherlands) electricity from windmills is also common. I do not know how this works in other parts of the world, but at least within the Nordic countries there is an open market for electricity trade. Consumers buy their electricity from a supplier, which normally also have its own hydro power plants. The suppliers are also allowed to buy electricity from other national or regional(e.g. Nordic) suppliers. While the electricity I bought from my supplier normally is clean (as in hydro power), every once in a while when they are low on water and buy for instance Danish electricity from coal fueled power plants or Swedish nuclear ones, I might be indirectly polluting. Hence, when I sign up for a given electrical supplier to give me my 220VAC, neither I nor they can guarantee what kind of power I'll end up with, not even nominally. Even if they could say that none of "their" electricity was polluting, the cables are free for all, and the electrons don't care where they came from, do they? So, in the elaboration of the depicted taxing schemes for electricity it would be necessary to either disregard the fact that there are different kinds of power plants, with different aspects to them, or it would be necessary to implement some kind of differential tax, based on how much "green", and how much "black" power you use.
  • by SEWilco ( 27983 ) on Thursday November 30, 2000 @10:40AM (#592404) Journal
    Well, you can get boards which use the laptop-designed mobile processors. An easy one to search for is "Tillamook" -- search for that along with words such as "motherboard", "VGA", "ISA", "PCI" (to filter out non-computer references).

    Some are single-board computers which may require that your computer case use PC/104 or passive ISA/PCI bus designs.

  • I'm serious, and no, this isn't some X-files nonsense. The correct term is "Reservoir-Induced
    Seismicity". A URL to get folks started:

    http://www.google.com/search?q=dams+cause+earthq ua kes&hl=en&lr=&safe=off

    Think of it this way. When you build a house, it settles. When you introduce a lot of weight anywhere onto the earth's surface, something has to deflect. In the case of softer ground, it settles, whilst bedrock deflects less, often imperceptably.

    The introduction of thousands of tons of weight into an area does cause problems. Everyone who's taken a geology class has heard how land uplifts after glaciers recede (weight removal), so why wouldn't land sink if weight is introduced?

    Two more examples. Quake Lake west of Yellowstone National Park. A large earthquake caused a mountain to fall, only a couple of years after a dam was built upstream. Hungry Horse Dam near Glacier National Park. Almost no siesmic activity, then a nice 7+ Richter quake only a couple of years after it filled.

    I know it sounds hokey, but honest folks. T'aint snake oil I'm selling here....
  • dhart:
    Energy efficiency is environmentally friendly...

    Energy efficiency is not intrinsically environmentally friendly. There is no shortage of clean energy that can be exploited in the form of nuclear fission power [stanford.edu]. California saves lots of energy by not using nuclear power to desalinate (make into fresh water) sea water. California alternatively drains Mono Lake and diverts [water-ed.org] water from the Colorado River and the San Joaquin Delta.

    Spending money to reduce energy consumption tends to hurt the environment by interfering with environmentally friendly factors such as economic and technological growth. It's a waste of engineering resources that could be put to positive environmental use -- such as devising cheaper and better nuclear power plants.
  • There's always a balance between damage from power consumption and pollution, which is often very hard to determine.

    I'd wager laptop-type components have a shorter lifespan. If this is true, then using such chips would mean more would have to be manufactured than regular ones.

    As to which is environmentally friendlier is anyone's guess.

  • Laptops general aren't made for constant use.
    The cooling on them isn't all that great, and hot components=higher failure rate.
    • Lower wattage CPU leads to lower heat dissipation
    • lower heat dissipation requires less machine cooling
    • lower machine cooling needs means lower room cooling needs
    Now, for some Apple hardware specs:

    G4 Cube - 225 Watts Maximum (Not sure why it's higher than the server...could be the video card.)
    Cube Specs [apple.com]

    G4 Server - 220 Watts Maximum (The G3 server is 170 Watts Max.)
    G4 Server Specs [apple.com]

    15" LCD Studio Display - 50 Watts Maximum (An equivalent CRT can almost triple this.)
    LCD Display Specs [apple.com]

    Those Wattage ratings aren't power supply output ratings. Those are what the device will pull from the wall, including any heat losses in the power supply.

    I can't find any other manufacturers who want to tell me their wattage requirements.


    --

  • I don't understand why you put the "unless it's a Powerbook" comment on only one of your criteria!

    * I heard the new Powerbooks in January are going to use PC133 memory. Powerbooks can now take one gigabyte of RAM, using 512MB sticks! Go see http://powerbookguy.com!

    * CPUs are as you said :)

    * Powerbooks are not slower except for possibly the PCI bus speed (I'm not sure); Apple uses virtually the same motherboard on all models, using their UMA (unified motherboard architecture).

    * Powerbooks have the same CPU's as desktops. I heard that the G3 cpu is the size of a person's thumbnail.

    * Installing LinuxPPC on my Powerbook (Wallstreet II series) was the single easiest installation of Linux I've ever done on any platform. It basically said "Oh, you have a Mac." It didn't ask for any hardware support and it was unconcerned with the fact that it's a Powerbook and not a desktop or anything else; they all use virtually identical hardware and it all just worked. Even at the time without a video driver, thanks to Open Firmware. It used SCSI, ethernet, the same video options as MacOS was in whenever booting Linux, PCMCIA, serial, and ADB. That's everything except video acceleration and sleeping, which came later.

    * You don't need PCMCIA RAID when you have Firewire or SCSI. You can use software RAID or an external hardware RAID controller. Older Powerbooks like mine have SCSI-II and newer ones have USB and Firewire.

    In short, there is no reason not to buy a Powerbook unless you want these:

    * A bare minimum lower price on an equivalent system if you're strapped for cash, but they're no more expensive on average than anything else, especially considering the mind-numbing amount of integrated features they have. In that case, I'd probably buy an older or cheaper Powerbook from somewhere like powerbookguy.com.

    * More than one CPU per host

    * More than one gigabyte of RAM per host

    * Some morbid juvenile fascination with native binary compatibility with IA32

    ===
  • Hydro is considered by environmentalists to be the single dirtiest [irn.org] power generation technology ever devised. Upstream from a dam, it displaces large areas of land for reservoirs and causes methane emissions (greenhouse gasses) from the rotting vegetation trapped under surface of the reservoir. Downstream, it causes massive ecological damage in the form of erosion caused by holding back sediment; changing river temperature; depleting groundwater; changing flow patterns (which threatens estuary ecology); and stopping flooding (necessary for healthy floodplain ecology).

    Fish can't swim upstream or downstream past dams.

    Hydro is also extremely dangerous. Large numbers of people are regularly killed [sabp.net] by dam failures, and there are two dams in the U.S. which could each nearly instantly kill 200,000 people.
  • rainer_d:
    The fossile energy-resources will be used up very quickly.

    "The world has about 1500 years of known coal resources at the current use rate. [ornl.gov]"

  • >I don't understand why you put the "unless it's a Powerbook" comment on only one of your criteria!

    ((In the voice of 'herbert' from 'Monty Python and the Holy Grail'))
    "Well, I'll tell you."

    I said "Unless it's a powerbook" only for one an important reason you overlooked:

    Powerbooks are (typically) THE MOST upgradable laptops in existance. Take ly 1400cs -- it was born with a 117mhz 603e. Right now, I could go buy a 333mhz G3 for it, and Sonnet is working on a 433mhz G4 upgrade for it. And take a Wallstreet or newer model -- it takes some work, but if history is any experience, we should be able to upgrade a stock 66mhz bus Wallstreet 233 to well over 600mhz with a simple card swap. Try that with most intel laptops.

    But anyway, on with some your points:

    * CPUs are as you said :)

    > * Powerbooks are not slower except for possibly the PCI bus speed (I'm not sure); Apple uses virtually the same motherboard on all models, using their UMA (unified motherboard architecture).

    This is true, except that there are power saving measures that are built into the system (Like CPU cycling, etc) and the fact that all parts in a laptop are designed for low power, not speed. It doesn't make them a whole bunch slower, but like 5-10% maybe. Using UMA board means their just fully compatible with each other, not the same.

    > * Powerbooks have the same CPU's as desktops. I heard that the G3 cpu is the size of a person's thumbnail.

    In some cases. I'm not sure how it is for the new Broze G3's, but on my old 1400cs/117, it *does* uses a special mobile 603e. Size has nothing to do with it, it's voltage and power usage.

    What I was saying was not for some kind of high-power server farm, but a medium load system. It you actually *need* a multi-firewire drive setup and 500mhz plus for a web site, just go buy a cube or a black&white G4.
  • jbert:
    Until we get lots of lovely clean (well, mostly) fusion power we have a problem.
    If we had lots of fusion power it would be a serious environmental problem since it would be so expensive (by all current estimates).

    ...I think power is underpriced.
    Power is vastly underproduced. Ecology demands that we produce exponentially more and that we do this with nuclear fission plants.
  • But the original story was about using laptop components. I took that to mean chips, drives, ram, etc. If you were suggesting just plugging up a laptop into a docking station and calling it a server (or workstation) I'd agree with you. But if someone were building a new hybrid (workstation with low power consumption drives, chips, etc.) there's no reason to believe that someone would put a fan on the chip. Also, if the goal of the original system is low power consumption, underclocking of some components could decrease the heat output, and lower consumption even more.

    Also, I'm not sure where you get that laptops are not made for constant use. Maybe they aren't. Again, I don't use 'em, and I don't buy 'em. But they are certainly sold and reviewed as such.

    Let me reiterate: there's a difference between throwing a laptop onto a desk and plugging in a monitor, and what the story is about. The story (in my estimation) is asking why not build one of those dinky PC's, but with low consumption, probably notebook based, parts.

  • by Xunker ( 6905 ) on Thursday November 30, 2000 @09:04AM (#592416) Homepage Journal

    Regarding servers: I was thinking myself this question a while ago, myself - the most logical choice being a laptop as the server.

    • Benefits:
    • Low Power
    • Very Low heat (little or no active cooling required)
    • variable power savng states (the machine can throttle it's processor usage dependent on how much it is being used).
    • Small (can fit in a closet)
    • Built-in battery backup
    • Can be found cheap, if you buy one with a broken screen. (just plug in an external monitor, and only turn it on when you need it)

    • Drawbacks
    • Sometimes are Proprietary (i.e. memory)
    • CPUs are Seldom upgradable (unless it's a Powerbook :) )
    • Mobile CPUs are usually slightly slower than desktop CPUs or the same MHZ
    • Mobile technology is usually 1-2 generations behind desktop machines (we have 1.2ghz processors on the desktop and 700mhz mobile CPUs have only recently appeared)
    • Software support may be lacking for some models (Especially Linux, BSD or NT4)
    • Expandability can be low - they don't make PCMCIA RAID controllers.

    In my case, I could probaly pick up a used p233 laptop with a busted screen for only a few hundred, buy two PCMCIA ethernet cards, drop Linux, *bsd (or even NT4 or Win2k) on it and it could do all my web serving and firewalling/NATing.
  • I'll say that the biggest issue I've had thusfar with a laptop server is the fact that it has a battery...period. A few weeks ago, I accidentally turned off the juice feeding the laptop, and I never knew it, because good 'ol Viggen was happily serving along, silently chugging on battery power. And of course, it all came to light later rather than sooner...
  • just buy cobalt. they give you shit servers which have nothing thermally active in em ... 300Mhz CPUs engineered to maximise profit for the company as embedded appliances selling for $1500-2000. they'll save you on electricity but you'll be paying thru your nose for the server.
  • The standard low-power-consumption server is the StrongARM powered Netwinder [rebel.com]. Originally developed by Corel and later sold to Rebel.com, the Netwinder is a very professional server. It comes with Linux and Apache installed and preconfigured and uses ~15 watts, less than most laptops.
  • Once The NetBSD Project [netbsd.org] finishes its Sega Dreamcast Port, and you can upgrade the modem in the Dreamcast to a network card, there will be no reason that you cannot use a dreamcast as a server (as far as I know). They are cheap (relative to the price of a computer) and the wow-factor of running a server off of a Dreamcast is way up there. Of course, network cards may never actually come and the full NetBSD port may never actually come, but I'd love to see this if it ever becomes possible.

    -----
  • You should look at experiments that the USGS did with injecting water into joints to determine if this could keep earthquakes from happening.
    I would have a hard time believing that the weight of lakes by itself could cause earthquakes.
  • Actually you can forget the display---all current Apple displays use the ADC connector which powers through the tower, not through a separate transformer in the display.

    ----
  • Rebel.com [rebel.com] make NetWinders [netwinder.org] which are StrongARM/Linux-based desktop and rackmount computers, intended as server or developer machines. They consume 15W and take up very little physical space.

    They include iRDA, speaker, mic, dual NIC, all the usual ports. The desktop takes up the same sort of size as a paperback book and the rackmount takes up half of a 1U chassis.

    The next generation of NetWinders will be based on Transmeta's Crusoe instead of StrongARM. Shame, because a NetWinder based on a 1GHz Intel XScale (successor to StrongARM) would have been very nice.

    I think that answers the question :-)

  • Due to having to be built to be dropped and exposed to greater heat/cold cycles, I would venture to say that the average laptop component is better built than a desktop. Further, if placed in a non-mobile, desktop environment, it should last longer.

    But I'm no engineer.

  • "The fossile energy-resources will be used up very quickly. Or at least those, that are easy to exploit."

    1. Coal is a fossil fuel (some has plant fossils).
    2. Oil and gas are geological fuels [cornell.edu], they're in places just being discovered now.
    3. Club of Rome [dieoff.com] 1972: World reserves 455 billion barrels. "We could use up all of the proven reserves of oil in the entire world by the end of the next decade," said President Jimmy Carter.
    4. 600 billion barrels used between 1970 and 1990.
    5. 1990: Oil reserves over 900 billion barrels [hiwaay.net].
    6. 1999: World oil reserves [doe.gov]: 1,000 billion barrels (5,000 of natural gas also).

    Of course, we're ignoring fission plants. Not that it matters when nobody is building power plants. Well, we can all fire up our backyard generators and see what that does to air quality.

  • ower seems to be cheap, a few pence/cents/whatever to keep lights running, run a computer, etc does not concern most (enough) people.

    Here in the state of California, we're rapidly getting away from this attitude, as our power rates have almost quadrupled since 'deregulation'. If this holds (there's immense political pressure coming to bear, as the politicians begin to realize that the largesse they've enjoyed due to a booming economy is completely dependent on business continuing to generate profits. What a radical concept. But I digress.), then the incentives to shave every kilowatthour are going to be very high. I can see low-power servers being of interest especially to the 'server-farm' businesses that have hundreds or thousands of computers in a single building. The lower the power requirements, the less expense, not only to power the servers themselves, but also for the air-conditioning, which can be just as costly.

  • I've been thinking about something like this for a while, and it's be a really neat thing to see big LCD screens, touchpads (or maybe touch screens, like some of the Fujitsu models [fujitsu-computers.com]), modular hot-swapable CD/DVD/Zip/Disk Drive/LS-120/Floppy Drive components and the support for PCMCIA cards with modems, network cards and such. Then when need to make a switch from working at the desktop at home to your mobile, almost all your components can go with you, including the data on the hot-swap drives. Another benifit would be very small size and lower power consumption, but all this is offsetted by a much higher cost than standard desktop components. Somebody out there has gotta be making these.

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

Working...