Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Regarding amps required for a server
New on LowEndTalk? Please Register and read our Community Rules.

Regarding amps required for a server

afnafn Member

Hi,

I am planning to do colocation for a server, but I am trying to figure out the hardware I need exactly, the budget and my colocation requirements.

In all colocation offers I see, they are billed by amps, 0.5 amps,1amp, 2 amps, etc.

Currently the way I estimate my requirements is by taking the CPU typical TDP and divide by voltage (220V for Europe).
So for example for an i9-9900 95W/220V ->0.43 So 0.5amps is enough
for an i9-10850k 125W/220V ->0.56 So 0.5amps is not enough, I need 1 amps

Is the way I do it correct? is there anything I am not taking into account about consumption of other components (the server itself and the drives)?

Thanks in advance for your help

«1

Comments

  • hzrhzr Member, Moderator

    ram, drives, fans, etc all use power

    you also need to typically buy more on top since you can only use 80%

    Thanked by 2jsg AlwaysSkint
  • afnafn Member
    edited April 17

    @hzr said:
    ram, drives, fans, etc all use power

    you also need to typically buy more on top since you can only use 80%

    Yeah, I was thinking about this too... So any idea how much more should I add to my estimates for these?

    So basically, to allow some margin for drives, fan, ram the only servers you can fit on 0.5 amps are 6 cores maximum, with TDP no more than 65w for the CPU to leave margin for the others?

  • @afn said:

    @hzr said:
    ram, drives, fans, etc all use power

    you also need to typically buy more on top since you can only use 80%

    Yeah, I was thinking about this too... So any idea how much more should I add to my estimates for these?

    So basically, to allow some margin for drives, fan, ram the only servers you can fit on 0.5 amps are 6 cores maximum, with TDP no more than 65w for the CPU to leave margin for the others?

    Also HDD will use more than an SSD; keep that in mind.

    Thanked by 1afn
  • DataIdeas-JoshDataIdeas-Josh Member, Provider

    @afn I use this to get a general idea of wattage used.
    https://outervision.com/power-supply-calculator

    Alien Data: VPS and Colo Based In Texas, Unmetered Bandwidth.
    RPIServers: Dedicated Micro Servers - Micro Colo - Unmeterd Bandwidth

  • JarryJarry Member

    These days TDP has little to do with real power consumption. Just check the latest Intel cpu: they can take twice more power than their TDP design value.

    There is no realiable way to estimate or calculate the power you need. The only proper way is to measure it. You can buy such a gear for a few bucks (check amazon for "power meter"), it is used between wall socket and power cord. Then you put your server under load (i.e. kernel kompile on all cores) and you get max power you need...

  • oplinkoplink Member, Provider
    edited April 17

    CPU watts is what uses the most power. Unless you have like 10+ non SSDs

    Every 120w is about 1AMP

    Do you have a target location (country) for your box?

  • afnafn Member
    edited April 17

    @DataIdeas-Josh Thanks! great link

    @Jarry, sadly this is not really possible in my case since I am actually trying to build a server (or maybe 2) such that it fits in 0.5amp and another one that has to fit in 1 amp. So basically I need to have a general idea of the hardware before ordering, can't wait till I get the hardware to measure...

    So far the general idea I have for my setup 2 DDR3/DDR4 ram, 2 SATA drives, 1 SSD, fan,
    with:

    • any CPU with tdp 65 or lower should fit fine in 0.5 amps
    • any CPU with TDP> 65 will need 1 amp

    I am guessing servers like the hetzner ex line
    https://www.hetzner.com/dedicated-rootserver/matrix-ex
    Should fit fine in 0.5 amps?

    @oplink,
    Thanks for the info, also for your question, yes NL or maybe FR

  • AlwaysSkintAlwaysSkint Member
    edited April 17

    @oplink said: Every 120w is about 1AMP

    For a nominal 120V (US) supply, yes, whereas it's normally 110V (EU) and in developed countries (:-p ) 240V. Notwithstanding supply voltage drops/spikes.

    Then=sequence/consequence; than=compare || Brought=bring; bought=buy
    Paid=NotSkint; payed=some naval thing

  • jsgjsg Member
    edited April 17
    • Easy way - if your main board IPMI (you are using a server board, right?) supports power adaptation/envelope then simply set it to max e.g. 110W.
    • Hard way - Find out the true avg and max power consumption of your processor (hint: AMD are more honest, intel basically provides a marketing number), your cooling (hint: 1 HU is wasting more usually), your memory, your disks, and of course your main board itself, plus even the power supply, which may waste between 3% and up to about 1/3rd, depending on quality, voltage (hint: 230V usually offers a bit better efficiency than 115V), and work load (hint: power supplies have more efficient and less efficient ranges)
      Also prefer SSDs over spindles where possible and 2.5" over 3.5" and be sure to get and look at the power consumption numbers of your disks; there can be quite significant differences. And forget about intel DT processors. Go for server processors with 45 - (max) 65W and it might work out.
    Thanked by 2Daniel15 AlwaysSkint

    The problem with democracy is that by definition > 85% of the voters are not particularly intelligent.

  • Server alone can't draw more than the power supply rating. ;) Though some are deliberately well over-specified.

    Thanked by 1TimboJones

    Then=sequence/consequence; than=compare || Brought=bring; bought=buy
    Paid=NotSkint; payed=some naval thing

  • dfroedfroe Member, Provider

    @AlwaysSkint said:

    @oplink said: Every 120w is about 1AMP

    For a nominal 120V (US) supply, yes, whereas it's normally 110V (EU) and in developed countries (:-p ) 240V. Notwithstanding supply voltage drops/spikes.

    And then there is Europe with its 230V. ;)

    Thanked by 1AlwaysSkint

    it-df.net: IT-Service David Froehlich | Individual network and hosting solutions | AS39083 | RIPE LIR services (IPv4, IPv6, ASN)

  • @afn said:
    @DataIdeas-Josh Thanks! great link

    @Jarry, sadly this is not really possible in my case since I am actually trying to build a server (or maybe 2) such that it fits in 0.5amp and another one that has to fit in 1 amp. So basically I need to have a general idea of the hardware before ordering, can't wait till I get the hardware to measure...

    So far the general idea I have for my setup 2 DDR3/DDR4 ram, 2 SATA drives, 1 SSD, fan,
    with:

    • any CPU with tdp 65 or lower should fit fine in 0.5 amps
    • any CPU with TDP> 65 will need 1 amp

    I am guessing servers like the hetzner ex line
    https://www.hetzner.com/dedicated-rootserver/matrix-ex
    Should fit fine in 0.5 amps?

    @oplink,
    Thanks for the info, also for your question, yes NL or maybe FR

    It doesn't sound like you actually have a special server or requirements that require Colo. Why even colocate instead of renting a dedicated server and limit your risk?

  • DataIdeas-JoshDataIdeas-Josh Member, Provider

    @TimboJones said:
    It doesn't sound like you actually have a special server or requirements that require Colo. Why even colocate instead of renting a dedicated server and limit your risk?

    Maybe because they want to colo.
    Some people don't like the idea of having their stuff on other peoples hardware....

    Alien Data: VPS and Colo Based In Texas, Unmetered Bandwidth.
    RPIServers: Dedicated Micro Servers - Micro Colo - Unmeterd Bandwidth

  • afnafn Member
    edited April 18

    @jsg thanks a lot for the hints.

    @TimboJones said:
    It doesn't sound like you actually have a special server or requirements that require Colo. Why even colocate instead of renting a dedicated server and limit your risk?

    1. Cuz why not :sweat_smile: wanna try something new.
    2. I actually do have a reason. I need a powerful intel CPU with integrated GPU that most providers won't offer in dedicated servers for cheap, for example i9-10900K with big storage. Right now, the cheapest that comes to my mind is Hetzner with i9-9900k (8 cores) for 69€/month+ at least 7-8€ for an SSD sor roughly 80€ per month for 8 cores i9, meanwhile I can get something slightly better with 10 cores in colo for 50€/month.

    Assuming the server costs around 800-1000€ Saving 30€ per month , means 2-3 years to get my money back, after it, you get a very cheap monthly price for a couple extra years remaining in the server's life. + The money spent on hardware is not lost money, you can always reuse/resell if you need an upgrade. Even if I assume my drives will die during the first 3-5 years, it is still almost the same price wise, but at least you get something slightly better for the money :) (+ you get the fun of it, the experience and the stress that comes with it xD, the full pack)

    Also to me, faster cpu = less computation time => time saved=> implicitly more money.
    Not directly, but in some sense there are some other hidden earnings if I have a better cpu. (and no, not using it for crypto mining)

  • DataIdeas-JoshDataIdeas-Josh Member, Provider

    @afn if interested I offer colo here in Texas.

    Alien Data: VPS and Colo Based In Texas, Unmetered Bandwidth.
    RPIServers: Dedicated Micro Servers - Micro Colo - Unmeterd Bandwidth

  • afnafn Member

    @DataIdeas-Josh said:
    @afn if interested I offer colo here in Texas.

    Thanks for the offer, I appreciate it :) I have been actually keeping an eye on your website since a while in case I need services in the US. But sadly I'm interested in Europe only (mainly NL, or maybe FR)

    Thanked by 1DataIdeas-Josh
  • The Supermicro 6039P-TXRT has 10A power supply.
    We just ask for one 20A circuit for every two servers.
    No need to measure anything or change any configuration.
    The power supply is not supposed to use more than its labeled value.

    The server actually has two power supplies, but works on just one.
    So it's calculated as 10A instead of 20A.

    Apparently the vendor puts in two power supplies, so that I can move the server onto another power circuit if the current one needs repairs.
    What a waste!
    Just checkpoint the experiment and turn off the server when you need to repair power circuits.

    The CDC says fully vaccinated people do not need iptables firewall and DDoS protection.

  • jsgjsg Member

    @yoursunny said:
    The Supermicro 6039P-TXRT has 10A power supply.
    We just ask for one 20A circuit for every two servers.
    No need to measure anything or change any configuration.
    The power supply is not supposed to use more than its labeled value.

    The server actually has two power supplies, but works on just one.
    So it's calculated as 10A instead of 20A.

    Apparently the vendor puts in two power supplies, so that I can move the server onto another power circuit if the current one needs repairs.
    What a waste!
    Just checkpoint the experiment and turn off the server when you need to repair power circuits.

    Uhm, that box is not available anymore ("status: discontinued") and highly likely uses (or wastes) way too much power, is clearly far beyond what @OP wants, and has > 1 kW power supplies, but supports no graphics (no built in GPU).

    Did you even read the OP?

    Side note: dual PSU is not a waste but meant to increase reliability/availability by keeping the server running in case one PSU or one power line fails. In fact I strongly suggest to insist on dual PSUs when renting a dedi (and they are standard except for the lowest end providers).

    Thanked by 1TimboJones

    The problem with democracy is that by definition > 85% of the voters are not particularly intelligent.

  • DataIdeas-JoshDataIdeas-Josh Member, Provider

    @jsg said:

    @yoursunny said:
    The Supermicro 6039P-TXRT has 10A power supply.
    We just ask for one 20A circuit for every two servers.
    No need to measure anything or change any configuration.
    The power supply is not supposed to use more than its labeled value.

    The server actually has two power supplies, but works on just one.
    So it's calculated as 10A instead of 20A.

    Apparently the vendor puts in two power supplies, so that I can move the server onto another power circuit if the current one needs repairs.
    What a waste!
    Just checkpoint the experiment and turn off the server when you need to repair power circuits.

    Uhm, that box is not available anymore ("status: discontinued") and highly likely uses (or wastes) way too much power, is clearly far beyond what @OP wants, and has > 1 kW power supplies, but supports no graphics (no built in GPU).

    Did you even read the OP?

    Side note: dual PSU is not a waste but meant to increase reliability/availability by keeping the server running in case one PSU or one power line fails. In fact I strongly suggest to insist on dual PSUs when renting a dedi (and they are standard except for the lowest end providers).

    To build on top of dual PSUs. It can be used for A/B power. In case you have a total power loss say from the grid. You have secondary power from battery/generator.

    Alien Data: VPS and Colo Based In Texas, Unmetered Bandwidth.
    RPIServers: Dedicated Micro Servers - Micro Colo - Unmeterd Bandwidth

  • jsgjsg Member

    @DataIdeas-Josh said:
    To build on top of dual PSUs. It can be used for A/B power.

    That's what I said (minus the 'can'. With me it is 'should').

    In case you have a total power loss say from the grid. You have secondary power from battery/generator.

    Nope, that's not how it's supposed to work. Both power feeds to the rack should be UPS'd. The point of a dual PSU is to continue running when either a power feed fails (or is in maint.) or a PSU fails.

    Thanked by 1TimboJones

    The problem with democracy is that by definition > 85% of the voters are not particularly intelligent.

  • DataIdeas-JoshDataIdeas-Josh Member, Provider

    @jsg said:

    @DataIdeas-Josh said:
    To build on top of dual PSUs. It can be used for A/B power.

    That's what I said (minus the 'can'. With me it is 'should').

    In case you have a total power loss say from the grid. You have secondary power from battery/generator.

    Nope, that's not how it's supposed to work. Both power feeds to the rack should be UPS'd. The point of a dual PSU is to continue running when either a power feed fails (or is in maint.) or a PSU fails.

    No, I agree that both feeds should be UPS'd but some places don't. 🤷
    And I was agreeing with what you said.

    Thanked by 1jsg

    Alien Data: VPS and Colo Based In Texas, Unmetered Bandwidth.
    RPIServers: Dedicated Micro Servers - Micro Colo - Unmeterd Bandwidth

  • user54321user54321 Member
    edited April 18

    @AlwaysSkint said:
    Server alone can't draw more than the power supply rating. ;) Though some are deliberately well over-specified.

    that is wrong. The OPP will trigger normally at ~130% of the rated power on the 12v rail, so depending on the efficiency of the PSU they will draw 135-150% of which the PSU is rated from the wall before they turn off.

  • DataIdeas-JoshDataIdeas-Josh Member, Provider

    @user54321 said:

    @AlwaysSkint said:
    Server alone can't draw more than the power supply rating. ;) Though some are deliberately well over-specified.

    that is wrong. The OPP will trigger normally at ~130% of the rated power on the 12v rail, so depending on the efficiency of the PSU they will draw 135-150% of which the PSU is rated from the wall before they turn off.

    Y'all people just asking for something to burn if pushing PSUs that far though.

    Alien Data: VPS and Colo Based In Texas, Unmetered Bandwidth.
    RPIServers: Dedicated Micro Servers - Micro Colo - Unmeterd Bandwidth

  • @DataIdeas-Josh said:

    @user54321 said:

    @AlwaysSkint said:
    Server alone can't draw more than the power supply rating. ;) Though some are deliberately well over-specified.

    that is wrong. The OPP will trigger normally at ~130% of the rated power on the 12v rail, so depending on the efficiency of the PSU they will draw 135-150% of which the PSU is rated from the wall before they turn off.

    Y'all people just asking for something to burn if pushing PSUs that far though.

    With all that power managment of modern CPUs and GPUs the usage is so fucking spicky that you can trip OCP or OPP even if your build is designed to only use 80% of your power budget. But beside that even if you push the PSU that hard it doesn't matter they will turn off before they would be overloaded. Of course some of the 10$ China ones won't but you can't get them legally anywhere anyways, so i hope nobody uses such crap.

  • @jsg said:

    @yoursunny said:
    The Supermicro 6039P-TXRT has 10A power supply.

    Uhm, that box is not available anymore ("status: discontinued") and highly likely uses (or wastes) way too much power, is clearly far beyond what @OP wants, and has > 1 kW power supplies, but supports no graphics (no built in GPU).

    Did you even read the OP?

    Yes, I did.
    I meant that, if OP wants 1A server, just take a PSU that is 1A.

    This model has no built-in GPU, but it has 11 PCIe slots (only four are x16 lanes).
    If you put 4x Tesla K80 in there, total power consumption is more than 1.5 kW, and you'll need both PSU.

    We added 3x ConnectX-5.
    They draw up to 27W each, so I guess the PSU only outputs 500W or so.

    I wouldn't mop over this model being discontinued.
    The next one, I'll ask for ConnectX-6 200Gbps and EPYC processor.


    @DataIdeas-Josh said:
    To build on top of dual PSUs. It can be used for A/B power.
    In case you have a total power loss say from the grid. You have secondary power from battery/generator.

    @jsg said:
    Nope, that's not how it's supposed to work. Both power feeds to the rack should be UPS'd. The point of a dual PSU is to continue running when either a power feed fails (or is in maint.) or a PSU fails.

    I could plug in two different power circuits, but then I would need twice as many power circuits.
    What a waste!

    It's simpler to just checkpoint the experiment and turn off the server.

    The only time(s) I used the second PSU is when I need to untangle the cords in order to move fibers/ other machines.
    In that case, I could plug in a second cord and unplug the first one.


    @user54321 said:

    @AlwaysSkint said:
    Server alone can't draw more than the power supply rating. ;) Though some are deliberately well over-specified.

    that is wrong. The OPP will trigger normally at ~130% of the rated power on the 12v rail, so depending on the efficiency of the PSU they will draw 135-150% of which the PSU is rated from the wall before they turn off.

    The PSU is labeled "980W" and "input 10A", so I suppose the inefficiency is already accounted for.

    The CDC says fully vaccinated people do not need iptables firewall and DDoS protection.

  • WilliamWilliam Member, Provider

    @afn said: So for example for an i9-9900 95W/220V ->0.43 So 0.5amps is enough

    In Power state 2. In Turbo (P1) a 9900K will use up to 170W for 25 seconds (across all cores), and up to 190W for 10 seconds on a single core.

    Intel TDP is about HALF of real usage in 9th/10th/11th Gen CPUs, as this was the only way to keep a marketing number somewhat near the cheaper, better, AMD CPUs.

    A i7-10700 (8c/16t) has a bench of 17400 at 65W TDP (P1 95W, P1+ 110W), 14nm Intel

    A Ryzen 5 5600X (6c/12t) has a bench of 22100 (30% more at 2 cores less) at 65W TDP (P1 65W, Turbo 90W), 7nm TSMC

    Intel also has no ECC support and no PCIe 4.0.

  • WilliamWilliam Member, Provider

    @user54321 said: that is wrong. The OPP will trigger normally at ~130% of the rated power on the 12v rail, so depending on the efficiency of the PSU they will draw 135-150% of which the PSU is rated from the wall before they turn off.

    The amp sticker has to account for loss and overload. The Amp stated on the sticker are a finite limit, and there is a legal requirement for them to be correct in EU.

    Also each PSU, at least HP/Dell/IBM, comes with a 10A+ glass fuse that limits it anyway.

    Thanked by 1AlwaysSkint
  • AlwaysSkintAlwaysSkint Member
    edited April 18

    @William said: The Amp stated on the sticker are a finite limit, and there is a legal requirement for them to be correct in EU

    This.
    The PSU output/efficiency and server internals are irrelevant, in my statement. If the PSU draws more than its' rated INPUT amperage, then it is faulty/out-of-spec/f'kin dangerous. Let's hope it blows a fuse!
    (OVH fire, cough)

    Thanked by 1yoursunny

    Then=sequence/consequence; than=compare || Brought=bring; bought=buy
    Paid=NotSkint; payed=some naval thing

  • WilliamWilliam Member, Provider
    edited April 18

    Notably at 210-240V (EU, Asia, anything aside US and Japan anyway) you should buy a 230V ONLY PSU (NOT A 80-280V universal) - Efficiency in Switch Mode Power Supplies (SMPS) is higher the more limited your input voltage is.

    Modern server PSUs run at 80-270V (80V is for Japans 100V system, 270V is for Eastern Europe where 250V+ is not rare) - Efficiency around 90-93% (VERY high, Gold+), Fixed voltage goes up to 95-98% (Platinum Plus).

    Also a universal 1000W PSU will come with a 10A fuse (120V=1200W, 220V=2200W) which is too much for EU, while the EU version will come with a 5A one (and a MOV) instead.

    Lastly, if you ever need it - All of this PSUs run perfectly fine on 90-230V DC - Chain 6 car batteries and your server will boot just fine. Polarity irrelevant (but should change it at times, to switch the full bridge rectifier to the other 2 diodes, DC will load only 2 always).

  • @William said: Chain 6 car batteries and your server will boot just fine.

    LOL. Just don't use a cheap Chinese car charger on them! (I've seen over 15V output from those time-bombs.)

    Then=sequence/consequence; than=compare || Brought=bring; bought=buy
    Paid=NotSkint; payed=some naval thing

  • @William said:

    @user54321 said: that is wrong. The OPP will trigger normally at ~130% of the rated power on the 12v rail, so depending on the efficiency of the PSU they will draw 135-150% of which the PSU is rated from the wall before they turn off.

    The amp sticker has to account for loss and overload. The Amp stated on the sticker are a finite limit, and there is a legal requirement for them to be correct in EU.

    Than it seems nobody does sell legal computer PSUs in Europe

    example



    Source: https://www.tweakpc.de/hardware/tests/netzteile/corsair_sf750_platinum/s08.php

  • WilliamWilliam Member, Provider
    edited April 18

    Yea, 14.8V should be maximum - My solar system has 58.4V set (/4 = 14.6V per battery)

    @user54321 said: Than it seems nobody does sell legal computer PSUs in Europe

    We talk servers here. Not off the shelf consumer parts.

    Also this PSU has a input rating of 240V*5A = 1200W (or 10A at 120V, also 1200W).

    The sticker is correct and the PSU legal.

    The DC voltage/Amps are irrelevant for regulators, only the wall draw matters.

  • AlwaysSkintAlwaysSkint Member
    edited April 18

    @user54321 said: Than it seems..

    Arrgh! Why do so many people have difficulty with this?

    @William said: The DC voltage/Amps are irrelevant for regulators, only the wall draw matters.

    I tried that one before but some folks just don't listen. ;)

    Then=sequence/consequence; than=compare || Brought=bring; bought=buy
    Paid=NotSkint; payed=some naval thing

  • @William said:
    Yea, 14.8V should be maximum - My solar system has 58.4V set (/4 = 14.6V per battery)

    @user54321 said: Than it seems nobody does sell legal computer PSUs in Europe

    We talk servers here. Not off the shelf consumer parts.

    Also this PSU has a input rating of 240V*5A = 1200W (or 10A at 120V, also 1200W).

    The sticker is correct and the PSU legal.

    If you go with that sticker and not the specified curents it tells you nothing, the PSU could draw at max 1200w or 300w because both will be labeled that way. So it tells you exactly nothing if you want to estimate the required amps for your colo.

    @AlwaysSkint said:

    @user54321 said: Than it seems..

    Arrgh! Why do so many people have difficulty with this?

    because reasons, I don't even understand the problem.

  • dfroedfroe Member, Provider

    @user54321 said:

    @AlwaysSkint said:

    @user54321 said: Than it seems..

    Arrgh! Why do so many people have difficulty with this?

    because reasons, I don't even understand the problem.

    than != then

    The one isn't just a different spelling of the other one.
    These are two different words with totally different meaning.
    Writing something else then [sic] you mean makes it more difficult for others to read.

    Thanked by 1AlwaysSkint

    it-df.net: IT-Service David Froehlich | Individual network and hosting solutions | AS39083 | RIPE LIR services (IPv4, IPv6, ASN)

  • AlwaysSkintAlwaysSkint Member
    edited April 18

    Let's just read the label correctly shall we?
    AC INPUT (the only relevant part in respect to this thread)
    100V @ 10A = 1kW
    240V @ 5A = 1.2kW
    So the PSU will (nominally) supply anything from 100 to 240V, with a maximum power draw of 1200W.
    If that server is off, uses nothing; if it's idle, who cares. Whilst running flat-out, the server may consume less but only reading real-time current draw will determine that. So the only thing to consider with near certainty is the maximum power draw, assuming a properly labelled, safe PSU. As @William said, put a 5A fuse in it and it can't draw much more - heck, put a 2A one in and listen to it pop when the server gets 'busy'. Good game! :)

    Done & dusted.

    If you want to determine the likely runtime of a UPS on the server, then that is entirely different.

    Then=sequence/consequence; than=compare || Brought=bring; bought=buy
    Paid=NotSkint; payed=some naval thing

    1. Don't forget that PSU isn't 100% efficient.
    2. The CPU isn't the only part of the server, you need to factor in EVERYTHING.

    My Blog rafalblog.xyz
    Contabo GmbH insanely cheap VPS |

  • 1. & 2. Still doesn't grasp it. :-|

    Thanked by 1skorupion

    Then=sequence/consequence; than=compare || Brought=bring; bought=buy
    Paid=NotSkint; payed=some naval thing

  • jsgjsg Member
    edited April 18

    Let's not get lost in over precision. After all input fuses tend to be +10% or 20% on a good day and quite a few major parts (like e.g. the tank inductors) are more like +-20% plus the ADC aren't really precise either, oh, and btw the input fuses usually have reaction times beyond the brown out stay alive times (typ. 10 - 20 ms) but they are meant to serve as worst case last resort (and legally required hence often just formally thrown in) devices anyway. And yes, of course they are utterly un-precise.

    @William said:
    Notably at 210-240V (EU, Asia, anything aside US and Japan anyway) you should buy a 230V ONLY PSU (NOT A 80-280V universal) - Efficiency in Switch Mode Power Supplies (SMPS) is higher the more limited your input voltage is.

    Largely theory (unfortunately).

    Lastly, if you ever need it - All of this PSUs run perfectly fine on 90-230V DC - Chain 6 car batteries and your server will boot just fine. Polarity irrelevant (but should change it at times, to switch the full bridge rectifier to the other 2 diodes, DC will load only 2 always).

    In emergency situations, OK, but generally I'd advise against that because if a device is specified for AC then it should be fed AC. One (of quite a few more) reason: in today's cost/profit optimized world most manufacturers will (ab)use input specs to save tenths of a penny here and there; condensators are a good example.

    That's also why, at least on the prosumer, not to even talk about the consumer segment you'll either not get or pay a leg and a kidney to get a 230V only input computer power supply. They didn't come up with 85 - 250V because they are friendly but because it allows for cheaper mass production, cheaper tested logos (e.g. CE testing) etc.

    And yes that also means that they often sh_t on legal requirements even of major markets like Europe. Example: I've seen quite a few computer power supplies (yes, even newer ones) that gobble up 10W and more in "off" state while the legally required max. is far lower.

    The problem with democracy is that by definition > 85% of the voters are not particularly intelligent.

  • WilliamWilliam Member, Provider

    @user54321 said: So it tells you exactly nothing if you want to estimate the required amps for your colo.

    It tells you the circuit size needed for full load and more relevant - the circuit breaker sizing.

  • AlwaysSkintAlwaysSkint Member
    edited April 18

    @jsg said: ..that gobble up 10W and more in "off" state..

    The perils of soft switching :( Where's that toggle switch?

    @jsg said: condensators

    That a predominantly American term? Have only heard of condensers before, to refer to capacitors. (I worked in a high voltage electrical lab where testing used to be done on capacitor paper. I used that same paper for other purposes.)

    Then=sequence/consequence; than=compare || Brought=bring; bought=buy
    Paid=NotSkint; payed=some naval thing

  • jsgjsg Member

    @AlwaysSkint said:

    @jsg said: ..that gobble up 10W and more in "off" state..

    The perils of soft switching :( Where's that toggle switch?

    Usually at the rear of the computer (where I guess 95% of people don't use it).

    @jsg said: condensators

    That a predominantly American term?

    Nuh, that's a jsg term, mea culpa. You are right, I should have said 'capacitor'.

    The problem with democracy is that by definition > 85% of the voters are not particularly intelligent.

  • DataIdeas-JoshDataIdeas-Josh Member, Provider

    @yoursunny said:

    @DataIdeas-Josh said:
    To build on top of dual PSUs. It can be used for A/B power.
    In case you have a total power loss say from the grid. You have secondary power from battery/generator.

    @jsg said:
    Nope, that's not how it's supposed to work. Both power feeds to the rack should be UPS'd. The point of a dual PSU is to continue running when either a power feed fails (or is in maint.) or a PSU fails.

    I could plug in two different power circuits, but then I would need twice as many power circuits.
    What a waste!

    It's simpler to just checkpoint the experiment and turn off the server.

    The only time(s) I used the second PSU is when I need to untangle the cords in order to move fibers/ other machines.
    In that case, I could plug in a second cord and unplug the first one.

    It might be more work and more cables however in a DC environment from what I've seen that is how most people run it.
    Granted you can do it via an ATS that works as well but IMO asking for more things to go wrong BUT if have those servers that only have one PSU it gives you the ability for redundant power feeds.
    Just really comes down to how mission critical is your equipment to stay online.

    Thanked by 1yoursunny

    Alien Data: VPS and Colo Based In Texas, Unmetered Bandwidth.
    RPIServers: Dedicated Micro Servers - Micro Colo - Unmeterd Bandwidth

  • AlwaysSkintAlwaysSkint Member
    edited April 18

    @jsg said: Where's that toggle switch?
    Usually at the rear of the computer (where I guess 95% of people don't use it).

    I was being facetious. ;) You're correct in the assessment, I'd imagine. :)
    (Need a long, thin arm to stretch around the back of a 1U or under a desk.)

    Thanked by 1jsg

    Then=sequence/consequence; than=compare || Brought=bring; bought=buy
    Paid=NotSkint; payed=some naval thing

  • @DataIdeas-Josh said:
    Just really comes down to how mission critical is your equipment to stay online.

    I run network emulations.
    If the machine is powered off suddenly, the last hour of experiment data is lost, but the experiment will rerun when it's powered on again.

    If I know when I need to unplug the machine, I'd schedule a shutdown earlier.
    It's not about experiment data, but to prevent filesystem corruption.
    A corrupted filesystem means I'll have to spend an afternoon in front of that huge air conditioner, reinstalling the machine.


    There's a mystery in one of the racks though.
    Every other Monday, between 07:30~07:35 UTC, all the machines in that rack would turn off at the same time.
    I found the electrician and they have no idea either.

    Then COVID hits and we had to move the machines to another location so that they are remotely accessible.
    I'd see whether the mystery shutdowns occur again when we get back later this year.

    The CDC says fully vaccinated people do not need iptables firewall and DDoS protection.

  • @William said:

    @user54321 said: So it tells you exactly nothing if you want to estimate the required amps for your colo.

    It tells you the circuit size needed for full load and more relevant - the circuit breaker sizing.

    :D
    Yeah sure do what ever you want mate it is your money that you throw out of the window for stupid over sized estimations.

  • WilliamWilliam Member, Provider
    edited April 19

    @user54321 said: Yeah sure do what ever you want mate it is your money that you throw out of the window for stupid over sized estimations.

    I see you never worked on a City DC buildout before, you use that high estimate because power density is pure gold in the future and you need it for HVAC sizing also... (+ less copper with 230V anyway, and no 80% rule in most of EU/Europe).

    EDIT: good example is the old Bucharest DC of Voxility (i think they own it or at least occupy much of it, old but ok), you crank that 32A up boiiiii, on up to 3 circuits, 22000W per rack, summer toasty ;)

    Thanked by 2jsg yoursunny
  • user54321user54321 Member
    edited April 19

    wtf you talk about? this thread is about estimating the power budget of a single server and now you come with buildout of a city DC? why not build your own power plant and plan the power grid for it? That must be very important too for your single server. Yeah sure that is all important but not for you with your single server

  • afnafn Member

    Wow, a lot happened here :o didn't expect my thread to bring all these comments...
    Thanks all for your contributions, I will make sure to read carefully the comments I missed :)

  • AlwaysSkintAlwaysSkint Member
    edited April 19

    @user54321 said: ..this thread is about estimating the power budget of a single server..

    @William 's point perhaps wisnae clear: people sometimes add things to servers at a later date, for example an additional disc - room for expansion. Hence, the recommendation/advice to use the rated PSU input.

    Thanked by 1yoursunny

    Then=sequence/consequence; than=compare || Brought=bring; bought=buy
    Paid=NotSkint; payed=some naval thing

Sign In or Register to comment.