Slashdot Mirror

Optimizing for low noise and cost based on off-the-shelf parts led me to this setup. It isn't the lowest power setup I can imagine, but noise and power are directly proportional in most cases. My goal was a very low noise, low cost always-on headless server running Linux with fault tolerant storage, at least one Gb NIC and enough processor and RAM to use for common development tasks. Based on measurements performed here and some guess work I estimate this is pulling 40W at idle and I can't hear it a beyond a meter.

MB: ASUS M2NPV-VM. AM2 socket with on-board Nvidia video, SATA etc. Not running a discrete video card is a large power savings.

CPU: Athlon 64 3500+ AM2 Lima core. This is a recent single core CPU from AMD. Easily obtainable from Newegg et al. The nice thing about it is the low TDP of 45W. This approaches portable CPUs while not costing so much. Stable at 1.2V (perhaps lower if I tried) and works well with cpufreq.

Case: Antec NSK3300 MicroATX. Small and quiet. Uses a high efficiency 300W power supply with a non-standard form factor. I doubt this machine can pull enough juice to get the fan moving at full rate. It's silent 99% of the time.

The rest: 1GB of "value ram", a pair of quiet 250GB WD disks and a Intel Gb PCI NIC I got somewhere. If you want to save more power run 1 disk, cut the RAM in half and don't add a fast NIC. Probably just under 30W at that point.

At that, throwing 8 more screws into the mix, and one more wire to plug-in isn't a big deal.

I know a whole lot of people who build their own computers, I only know a handful, myself included, who are willing to mod anything, and touching PSUs isn't my favorite pastime, either. If you think modding hardware is not a big deal to most people, you're a bit out of touch. A long time ago, when I built my first two computers, I was scared shitless - having saved up for the hardware and all; I wouldn't have dared opening up the PSU. I'm not saying it can't be done, I'm saying it's a hassle, and I'd rather buy a PSU that's quiet from stock; especially since these days I can afford the 10 extra bucks for a nice PSU if it saves me an hour of work.

I'm not sure what gives you that idea. They very clearly rate notebook hard drives as having much, much higher vibrations than any desktop hard drives.

Ehm. No. Their vibration scale is x out of 10, 10 being completely without vibrations. Laptop drives routinely get 9/10, 3.5" HDDs vary between 3 and 6. "We've been recommending 2.5" notebook drives for desktop systems since early 2004." Why? First among other things: "Much lower noise."

FWIW AAM on or off doesn't make much of a difference with my drive, especially since I suspended it.

At least from my experience with my Mac mini. Even with Eclipse running and compiling things in the background, my power usage has never climbed above 60 watts.

SilentPCReview did some tests with iMacs as well. They loved them (no wonder IMHO). They did a 2xCPUBurn+ATI tool und XP to stress both cores and graphic card and never got more than 73 watts power usage even WITH the LCD on (link).

The iMac 24" used a maximum of 138 watts (link).

Bye egghat.

At least from my experience with my Mac mini. Even with Eclipse running and compiling things in the background, my power usage has never climbed above 60 watts.

SilentPCReview did some tests with iMacs as well. They loved them (no wonder IMHO). They did a 2xCPUBurn+ATI tool und XP to stress both cores and graphic card and never got more than 73 watts power usage even WITH the LCD on (link).

The iMac 24" used a maximum of 138 watts (link).

Bye egghat.

And the lesson is to use mobile and integrated components, in laptop form or not. Here is a review of the power consumption of an Apple iMac, but you could easily build a computer with a laptop CPU/GPU etc.

An added benefit of mobile components is that they don't need much cooling, so you get a quiet computer, too.

That's not at all surprising, but not for the reason you think. Laptops work by using a battery to moderate the power consumption. You drain the battery down to typically 95% or so before your charge circuit kicks in and brings it up. That means that unless you measure over a long period of time, you will get a false low reading because the external brick is in trickle power mode.

Even if you measure over a long period of time, however, a laptop will still always be more efficient than a desktop for a number of reasons:

• Smaller display = less power, generally speaking. They also usually don't have as bright a backlight, which also generally means less power.
• 2.5" hard drives use dramatically less power than 3.5" hard drives
• Laptops on the average have less RAM, and that's a huge power sink
• Laptop drivers generally tune CPU use in a more conservative way to favor lower power over better performance
• Laptops have a lot less hardware. Most of the legacy I/O isn't there, the built-in keyboard and trackpad don't draw nearly the power of an external keyboard and all the required support hardware, etc.
• The battery averages out power drain, which means that the power supply's peak power doesn't need to be significantly above its continuous power. This allows you to use a smaller supply, which results in significantly reduced waste.

That last one bears restating. While it is true that switching supplies do draw power from the mains that is somewhat proportional to the current drain on the output, they are most definitely not linear. This means that efficiency for smaller supplies is significantly better than for larger supplies. While you don't want to undersize a supply, oversizing the supply will result in fairly significant power waste. By being able to ignore the need for extra peak power (because of the battery), you would expect the efficiency of a laptop power supply to be several percent better (on the average) than a desktop even if all other factors were equal.

The only real question, then, is why desktops don't all have built-in backup batteries in them. It would be far more efficient than a UPS, and it has benefits in terms of power supply efficiency as well....

Google didn't see overtemp failures only because google kept their drives cool. Possibly too cold. Their graph
cuts off at only 10 degrees hotter than a typical PC. But if you extrapolate the data on the right hand side of
the graph, you see that drives fail at higher temperatures just as expected. Also, they appear to have looked
at average temperatures over the life of the drive, not the temperatures near the time of failure. And they
totally ignored temperature fluctuations.

In fact, the conclusion one should reasonably draw from their data (if it can be trusted and I called that into
question saturday) is that drives are designed to operate at 40 degrees C (which, happens to be the operating
temperature of the hard drive on this machine right now in a typical mid tower case) and that any deviation higher
or lower will result in increased failure rates:

But it is also possible that the cooling systems, and not the temperatures themselves, are possible for the
drive failures seen in googles systems. They had some hard drives (the ones particularly responsible
for the low temp failures) that were operating at around room temperature. With light fan cooling, a drive
operates at around 20 celsius degrees above ambient. So how do you get an operating temperature around
room temperature? You cool the server room to freezing, you put A/C evaporator coils inside the server boxes or racks,
you water cool them, or you sandblast the drives with hurricane force winds (slight exaggeration). All of those
approaches raise the possibility of creating environmental hazards other than temperature.

But it is quite possible that it is just the temperature and that drive manufacturers have done the sensible thing and optimized their designs for the typical operating temperature of a drive. I also point out that there are
a number of failure modes associated with over-temp, under-temp, and temperature variation.

In a typical PC, the most likely cause of an overtemp failure is a fan failure.

http://hardware.slashdot.org/comments.pl?sid=22297 8&cid=18063644

Using google, ironically, I found at least one example dating back to 2003 of people discussing the effects of
too low an operating temperature (i.e. room temperature) through excessive cooling adversely affecting hard drives (not even getting into industrial or outdoor temperature ranges). And I wasn't even looking for that: http://www.silentpcreview.com/forums/viewtopic.php ?t=7677

Conclusion: For best results use a closed loop temperature control system with redundant variable speed fans to keep
the drive itself (not the ambient air) at a constant temperature of 40 degrees C. Or operate your machine with
moderate cooling in an environment comfortable for humans and use software to power down the drive and raise alarms if
it gets much above 50 degrees C. Whether you should shut down if the drive gets below 25 degrees C (after time to
come up to operating temp) is debatable. If you have had a major heating system failure or a broken window in winter, the drives own heat might be giving it some protection but the drive is also more vulnerable when operating than when shut down.

Your link is to TDPs. Which are not representative for normal power usage, and hence pretty useless. You need to look at figures at http://www.silentpcreview.com/ or a similar place. Last I looked, the desktop Core 2 Duo is more efficient under load with Athlon 64 X2 being more efficient in idle (apparently because Intel's EIST is not really working).

I am considering the 17" iMac, which is basically a laptop on a stand. SilentPCReview did a review of the older Core Duo (not Core 2 Duo) model and the most wattage they could get it to pull was 73 watts, and that was running Windows with 2 x CPUBurn and a ATI Tool stressing the video card.

http://www.silentpcreview.com/article594-page1.htm l

Another option might be to build a PC with AMD's new 65nm Athlon X2 CPUs, a motherboard with a built in GPU, a 7200rpm laptop drive, and LCD screen. You'd think that would be fairly miserly.

Nice data. I'm surprised to see the Antec PSU more efficient than the Seasonic as Seasonic is generally regarded as one of if not the most efficient PSU manufactures. In fact, Seasonic manufactures many PSUs for other companies, Antec included so I would not be surprised if the Antec PSU you tested was in fact a rebadged Seasonic unit.

There is a full review of the PSU at SilentPCReview who has many full reviews of PSUs including efficiency tests.

In general, any PSU with active PFC will generally pretty efficient, especially compared to any PSU without active PFC. An easy way to tell if a PSU has active PFC is to see if it can take 100-240v AC without a switch.

Nice data. I'm surprised to see the Antec PSU more efficient than the Seasonic as Seasonic is generally regarded as one of if not the most efficient PSU manufactures. In fact, Seasonic manufactures many PSUs for other companies, Antec included so I would not be surprised if the Antec PSU you tested was in fact a rebadged Seasonic unit.

There is a full review of the PSU at SilentPCReview who has many full reviews of PSUs including efficiency tests.

In general, any PSU with active PFC will generally pretty efficient, especially compared to any PSU without active PFC. An easy way to tell if a PSU has active PFC is to see if it can take 100-240v AC without a switch.

Quiet? The typical "quiet" fan sold these days is about 20-25dB. Many are higher despite the bogus advertising that reads "30dB is a quiet library". Unless you're playing music, have the TV on, or live beside a trafficked roadway, that translates in a fairly noticeable whine. Multiply that by any number of fans per system, and you end up with a lot of unwelcome noise. And there is no indication that plastic fans are going to ever be any quieter, or less numerous, so "these days" isn't very different than "those days" and will be the same well into the future.

I needed more space so I buy the then top-of-the-line Seagate 300GB 7200.8. I believe this was the first to use Perpendicular Recording Technology.

http://www.silentpcreview.com/

This site has been very helpful for me.

I'd agree with this, Alienware PCs/Notebooks look fairly atrocious. I'd much rather have a nice small Shuttle (I currently own a SN45G, bit old, but still very pretty), or Antec if I need the extra space. And while my Dell laptop isn't "pretty" per-se, it isn't garesh either.

A cheaper 80+ 250W PSU would cost a little under $40 before taxes. (I think 80+ is the new buzzword for 80% or more efficient PSUs). Older PSUs, say 2+ years, were typically in the 70% efficiency range. There are a bunch of articles at http://silentpcreview.com/ and other sites about this sort of thing.

I ran the numbers a while ago for one of the PC's around here. The last time I ran the calculations, it costs around $50/year to power that PC for about 6 hours per day. So the break even for me is somewhere around 5-8 years! So while the power grid would get a bit of a break, financially, I wouldn't.

I still might get a new PSU, but that's more because I have serious doubts about the quality of the power coming out of the current one (a suspicious # of hard drive deaths...) but that's a separate issue.

For new purchases, definitely go for the more efficient PSUs... as far as "upgarding" goes... it's borderline at best, at least for me.

cliffski wrote:

And why does my PC have such a ludricous power supply anyway, especially when im just surfing, do I really need it all?

From http://www.silentpcreview.com/article28-page3.html :

Without getting into technical details, the nature of a switching power supply is that it delivers as much power as is demanded by the components. This means that when installed in a PC whose components require 200W, a 400W PSU and a 250W PSU will each deliver 200W. ... As long as there is adequate power, higher efficiency is the key to cooler, quieter PSU operation.

In other words, power supplies only draw as much power as they need to meet the computer's power needs, so all power supplies will draw less power when you're web surfing than when you're playing games or running intensive calculations. What's most important is how efficiently a power supply converts AC power into DC power, and hence how much AC power it must draw to supply that demand.

The only thing a computer case is conducting is sound, the heat is removed by air convection(or alternatively water convection can do most, but not all of the work), that's why wood or acrylic are actually more efficient materials for a case to be made out of in terms of cooling per noise, see: http://forums.silentpcreview.com/viewtopic.php?p=1 00913#100913

The claim that a case needs to be made out of a conductive material is basically a hoax perpetuated to sell aluminum cases.

There's already a small-scale example of this: the PicoPSU. You use an external 12V power brick, and then internally replace your entire computer PSU with something about the size of a matchbox. However, it is only 120W, and a bit short on connectors.

You can buy a ps that takes 12 volts and turns it into the other voltages. It is tiny and isn't much larger than the power connector on the motherboard. It isn't even that expensive. People commonly use them to make computers that are small enough to fit in whisky bottles or other interesting enclosures. :-) I've been using such a ps for a while and it works well.
http://www.silentpcreview.com/article601-page1.htm l

As for CnQ, I haven't been able to measure any difference in idle power usage whether it's on or off.

I've actually been trying to spec. out a PC based PVR the last few days -- my total pretax cost is coming around to CAD 450.

It should be relatively easy to get your power usage to under 150W at load, and half of that at idle. It's still a lot more than a DVD player, but certainly not 350W. The trick is to be selective about components -- you don't need the highest end, most power consuming, hardware for this. Here's a link to an article at Silent PC Review on these things -- note, the system they've tested does not include a TV tuner, and uses a laptop harddrive, so figure on them being on the low side.

Don't forget, after you've built a PC based PVR, you get a whole lot more flexability that you could possibly get from a standalone device. Imagine 3 years down the road you want to switch to HDTV, use next-gen DVD's, or get multiple TV tuners? Well, each of those will be a simple $100-200 dollar upgrade.

By following this guide, http://www.silentpcreview.com/article16-page1.html , I have built two machines identical to their setup and modded my power supply cpu with a low rpm fan. These machines have run solid for almost 4 years, until one of my motherboards crapped out. They are quietest piece of hardware I have owned since I sleep right next to them.

1. The power supply is 14dbA? Yea right. The reviewers didn't even test it. That number is highly unlikely, especially from an unbranded PSU.

2. What about case fans? What's the dBA on those? What about the CPU fan, which isn't even included? Both of these will have a big impact on the total overall noise of the system.

3. Not to mention the hard drive! But enough.

This is not a "silent" HTPC. It's a quickstart HTPC. It should be judged on those grounds. Calling it truly silent is just going to confuse people.

See http://www.silentpcreview.com/ for real silent computing.

Be aware that the power usage of the old Athlon 900 and current Athlons is stated differently.

The old, hot chip may use 60W constantly. A new rating (TDP) of 65W is shared amongst a range of current Athlons of different speeds and represents the maximum heat dissipation possible for all of them so that manufacturers know how to design and size cooling solutions. A new Athlon processor will use much less power when idle or not fully busy and may never reach the rated number with a real workload, especially if it's at the lower end of the range.

Intel rates its chips yet another way, by measuring the power used when running a certain workload, not the maximum dissipation possible by any workload. This means that Intel's numbers are understated compared with AMD's. The article below shows how usage can be 10W or more higher than the "typical" rating. I think the desktop Core chips may also be able to reduce power usage when idle though.

http://www.silentpcreview.com/article169-page3.htm l

This drive has also been reviewed by SilentPCReview for those of us, who are more interested in noise than in performance. Another 500 GB Caviar model, the WD5000YS, was covered by StorageReview -- IIRC, the differences between those two drives are in the firmware.

http://www.silentpcreview.com/

http://www.silentpcreview.com/ A few thousand people there seem to think noise is more important than performance.

I was going to say that.

One good place to look for people's experiences with these cards is the Silent PC Review GPU forum.

But I understood that for TDP AMD quote worst case, but Intel quote "typical".
http://www.silentpcreview.com/article169-page3.htm l

Don't feel sorry for me, there is no point.

I bought a S754 system almost 2y ago. You can still buy new processors for my very own mobo today, great ones too, in the form of the latest and greatest Turion processors. A year later I bought the almost exact same system for my parents.

Do you know any current mobo for which you think you'll be able to buy a compatible CPU in 2 year's time ?

Also it was incredibly cheap, the CPU works with passive solid-state cooling and is still running like a champ. It was thus because AMD was pushing 939 instead, but I couldn't see the benefits at the time, and still don't except perhaps for PCIe, which I have no use for right now. At any rate I couldn't have bought AMD + PCIe then, and 939 was much more expensive, it was really an easy decision.

I'm a regular visitor to Silent PC Review (SPCR) where they talk about CPU heatsinks at great length. Only one of the ones in this article (Thermaltake Sonic Tower) do I ever recall seeing mention of at SPCR. The noise levels they're talking about (~45 dBA) are just way over the top from my point of view - I'd be reluctant to consider anything over about 25 dBA, which is about 50-100 times quieter.

If you're a mad overclocker who plays FPS games with sound through your stereo system with volume on 11, this is a useful review. If you want quiet, go to SPCR.

I use a Scythe Ninja passive (fanless) heatsink. Until about 10 days ago, I had a nearly inaudible, single fan system. (I upgraded my video card to be able to play Oblivion, and I'm waiting a few months for it to fail before I void warantee by replacing the active heatsink it came with.)

I was looking for a good heatsink/fan for my dad's Intel P4 560 (he does some gaming) and I settled on the Scythe Ninja and a Nexus 120mm fan. The fan itself is extremely quiet (you can hardly hear it at all) and it keeps the CPU very cool; if I remember correctly, it was hovering around 50C during the stress tests.

A great site to check out, which I based my purchase on the reviews from, is SilentPC Review. It has a lot of in depth reviews and what not on cooling devices.

Us silent pc enthusiasts have known about this for a long time.
I highly recommend this site: http://www.silentpcreview.com/

(I don't work for them)

How important is heat, really?

Heat is a huge consideration to many people, often the deciding factor.

Assuming that the machine has been engineered sufficiently well to prevent the processor from melting down

It doesn't matter how well the machine is engineered. If you have hot componentry you'll have a hard time getting rid of the heat without making a lot of noise, especially under load.

But I never even considered not buying one because of the heat

What choice did you have? With laptops (especially Apple) you basically take what you can get. There's very little mention of heat or cooling considerations at all.

And no consideration at all to desktop buyers

I bought an Athlon X2 solely because it runs much cooler than the P4.

and in server rooms where it is a consideration... they'll have an A/C system anyway
The consideration is power consumption. More heat means more power draw means more expensive.

I doubt Intel is going to lose any customers because their chip gets too hot.

They lost me in the last round. Thankfully they're finally about to put the P4 to rest and we can get back to the good old P3. I mean, P-M. I mean, 'Core'. Whatever.

By the way, once you start caring about heat (and you will!) go here for starters: http://www.silentpcreview.com/

And by the way - this topic sounds very exciting, but lots of people are running myth on quiet EPIA mobos. Check out mythtv's hardware database. Its not new or even hard.

1. Buy mobo, case, HDD, encoder card, DVD. Or buy one of these pre-built ones
2. Plug it all in.
3. Insert knoppmyth CD.
4. install
5. ???
6. profit... errr... I mean watch TV.

For anyone here looking for PC silencing information, head to http://www.silentpcreview.com/, it's a great resource. I silenced my AMD64 system last year using their product reviews and help from the enthusiasts on the forums.

The only thing I have to say about this is:

http://www.silentpcreview.com/

I'm stunned it hasn't been mentioned already. It's pretty much the only place that you can take their computer noise information seriously and has a great community as well. Everything from very reliably sound tested retail parts to crazy do-it-yourself projects (including shoving it in a cupboard) are intelligently talked about there. I know this sounds like an ad but go there yourself, you'll see.

It's much worse than that! AMD's Thermal Design Power is set higherthan the Maximum power used (worst case scenario), whereas Intel's refers to the Average power. This can translate into a difference of approx 50 watts.

Examples of the actual power consumption differences:
http://www.tomshardware.com/2005/05/09/amd/page20. html

Here is a detailed discussion:
http://www.silentpcreview.com/article169-page3.htm l

If they made their processors slower, then they would be "innovating"? What do you want processors to do, really?

Lower power consumption? Lower heat output to the point where I don't need a Rube Goldberg Device?

and under Windows, I'm quite happy with crystalcpuid http://crystalmark.info/?lang=en

It also happened that my athlon64 3000+ is still stable when going under the default lowest consumption state (1GHz at 1.1V). It works fine at 800MHz and 0.8V

a nice guide here: http://www.silentpcreview.com/article231-page1.htm l

This other SPCR article shows that your computer probably draws around ~250W at full load. There are very few situations today where you need a 500W or a 600W power supply.

How about a stack of twenty PIII 500s with CPUs from eBay for thirty bucks in shipping. File servers don't need Ghz plus CPUs. Slot 1 PIIIs are fanless too. If I recall the PIIIs are about thirty watts which is actually less than the PIIs.

Here's a CPU wattage comparison list.

http://www.silentpcreview.com/article31-page1.html

For my home systems in the computer room, I'm replacing each to use the Seasonic S12 430's -- they're great. I agree, they can't be heard over the the cpu/case fans and hard drives.

I have the Silverstone in a HTPC in the livingroom. Booting from the network, it's silent. Seems almost unnatural. I wanted a "standard" PSU and this fit the bill great. It's load is moderate and is never more than just warm to the touch.

Both have high Power Factor Correciton so my UPS's seem to last longer (<note to="self">I'll have to measure that and see.</note>) I found SilentPCReview.com of the Seasonic S12 430 and Silverstone ST30NF interesting and more in depth. Specifcally, how the Silverstone's rating is when used at 200~240VAC.

I've ordered four or five from case-mod.com over the past six months.

For my home systems in the computer room, I'm replacing each to use the Seasonic S12 430's -- they're great. I agree, they can't be heard over the the cpu/case fans and hard drives.

I have the Silverstone in a HTPC in the livingroom. Booting from the network, it's silent. Seems almost unnatural. I wanted a "standard" PSU and this fit the bill great. It's load is moderate and is never more than just warm to the touch.

Both have high Power Factor Correciton so my UPS's seem to last longer (<note to="self">I'll have to measure that and see.</note>) I found SilentPCReview.com of the Seasonic S12 430 and Silverstone ST30NF interesting and more in depth. Specifcally, how the Silverstone's rating is when used at 200~240VAC.

I've ordered four or five from case-mod.com over the past six months.

God knows why I put https in the link - correct link is SPCR. So much for using preview feature :(.

SPCR generally does rather high-quality reviews against consistent baseline, which allows to do the comparison. It also covers all aspects of "silent computing" (and is essentially dedicated to it - including forums). For this particular PSU, just go here.

SPCR generally does rather high-quality reviews against consistent baseline, which allows to do the comparison. It also covers all aspects of "silent computing" (and is essentially dedicated to it - including forums). For this particular PSU, just go here.

the NeoHE supplies didn't seem to fare all that well when reviewed, some seemed incompatible with Asus nVidia boards, and there seem to be reliability issue ...

http://www.silentpcreview.com/article273-page5.htm l

I've been undervolting my HP zv5000z Athlon 64 notebook since July of 2004, and I got the idea from Silent PC Review where people were undervolting desktop Athlon 64s. I switched to using CrystalCPUID to manage speed and voltage since that initial post was written and most people have switched to RMclock. Lots of people on R3000 Forums and the HP forum at NotebookReview.com have been undervolting their notebooks. I believe I was the instigator on both of those forums.

Anyhow, I'm still undervolting, now with a Mobile-class Athlon 64 3200+ CPU that I swapped in some time ago (HP only used DTR-class chips). AMD is very conservative with their default voltage levels so there's lots of room to work with. I've set my CPU to run at roughly Low Voltage-class levels, close to what a Turion ML is rated to do. The machine is solid. I got great battery life before and even better battery life now. I was able to play DVD video for 3 hours (12 cell battery) and get closer to 4 hours in general use. That's with a 7200RPM HD too. Not bad for a desktop-replacement behemoth.

Note that if your machine usually runs at AMD's 800MHz idle speed, undervolting won't buy you much (if anything). AMD's PowerNOW! is already extremely efficient in normal use. If it jumps to full speed a lot, undervolting does wonders.

So, there's no need to suffer with a 32-bit-only Intel CPU and their awful integrated GPU if you want a long-lasting notebook.

(Anyone know how to undervolt under Linux? That's the one thing I've been missing...)