LOW Power PC for light aplications.
Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee
-
- Posts: 34
- Joined: Sat Apr 23, 2005 4:33 am
- Location: Sweden
- Contact:
LOW Power PC for light aplications.
Hi,
I’m thinking of building a LOW power, low noise, always on pc for the following application.
Light server usage. (File sharing with family and when I’m traveling)
Woip client (replacing my home phone)
Internet and office applications such as mail, calendar and text editing.
Some home automation.
Music player over S/PDIF
DVD burning.
I need help finding a suitable motherboard and processor.
I’m thinking about going for a VIA EPIA micro-ATX but I’m open to any suggestion. Or should I go for a cheap laptop?
Any other sugestions for a system like this?
I’m thinking of building a LOW power, low noise, always on pc for the following application.
Light server usage. (File sharing with family and when I’m traveling)
Woip client (replacing my home phone)
Internet and office applications such as mail, calendar and text editing.
Some home automation.
Music player over S/PDIF
DVD burning.
I need help finding a suitable motherboard and processor.
I’m thinking about going for a VIA EPIA micro-ATX but I’m open to any suggestion. Or should I go for a cheap laptop?
Any other sugestions for a system like this?
-
- Posts: 34
- Joined: Sat Apr 23, 2005 4:33 am
- Location: Sweden
- Contact:
No, that's Pentium 4, Prescott core to be more spesific. As long as you get a P3 with a Coppermine core you'll be fine. Even a Coppermine Celeron would probably be adequate, and they have TDP of starting from ~10 watts.
http://en.wikipedia.org/wiki/List_of_In ... 8180_nm.29
http://en.wikipedia.org/wiki/List_of_In ... 8180_nm.29
-
- Posts: 1608
- Joined: Tue Jan 04, 2005 4:02 pm
- Location: United States
I had an old Compaq P2 350MHz box that was really nice. CPU was passively cooled, and the system had two quiet (almost inaudible) 80mm fans, one blowing into the case and over the the CPU heatsink, and one in the PSU. It's Achilles heel was the hard drive, those old ones whine like crazy. That's easily enough replaced, though, and since you plan to use this for a file server I'd assume you'd be getting something bigger anyways.
I've messed around with a couple P3 systems and they were all screamers.
I've messed around with a couple P3 systems and they were all screamers.
-
- Posts: 312
- Joined: Tue Jul 24, 2007 3:57 pm
- Location: Minnesota
Via Epia should work fine for everything up there, and is low power. I'm not sure about DVD burning; that can be some-what CPU intensive, I think.
A AMD 45W/motherboard with onboard/memory would also be acceptable, faster, and probably as inexpensive.
A Mac Mini would also probably work fine for this as well.
A AMD 45W/motherboard with onboard/memory would also be acceptable, faster, and probably as inexpensive.
A Mac Mini would also probably work fine for this as well.
-
- Posts: 34
- Joined: Sat Apr 23, 2005 4:33 am
- Location: Sweden
- Contact:
-
- Posts: 2
- Joined: Thu Nov 29, 2007 6:25 am
Re: LOW Power PC for light aplications.
It seems like most of the other people who responded focused on your list of applications and not on your low power and low noise specification.prostetnic wrote: I’m thinking of building a LOW power, low noise, always on pc for the following application.
I once tried to get an old PIII desktop set up as an always-on computer but the CPU fan was loud as heck and the whole system drew somewhere around 45 watts, which to me was too much.
I have a VIA Epia SP8000E fanless computer that I have on 24-7 as a MythTV box, and I believe it draws around 15-18 watts not counting the PVR-150 capture card, but including a 2.5 inch 80 gig HD. I have not tried to really use it much for anything else beyond MythTV, but I suspect it could probably handle your applications, and a more powerful Via board should definitely be able to do all you mention (though it may draw more watts than mine).
My other always-on computer is an circa 2001 laptop (don't remember the processor) running Ubuntu command-line version (no GUI) which draws about 14 watts with the LCD backlight off. Its basically just a home LAMP server plus print server.
After doing a fair amount of research on low power motherboards my conclusion is that if its low watts that you care about most an old laptop with USB 2.0 ports is the best bet for an always-on computer that sees light usage. However, if you are content with a system that pulls in the 25-40 watt range then VIA Mini-ITX boards, or the new Intel board (forget its name) are more fun and have more capabilities and expandibility.
-
- Posts: 62
- Joined: Tue Feb 20, 2007 2:01 pm
First question is why the power target? If for environmental reasons - consider the power to make/ship/etc. the components as well.
(May be greener to reuse existing than to make new - typically much of power used by a computer goes in to the manufacture.)
If want to cut power/noise on a P3 - what about using a picoPSU? I haven't tried one, but review said it runs more efficiently than many on a low power system. (If you can use the the 80w brick, it is fanless - so probably prety quiet too.)
(May be greener to reuse existing than to make new - typically much of power used by a computer goes in to the manufacture.)
If want to cut power/noise on a P3 - what about using a picoPSU? I haven't tried one, but review said it runs more efficiently than many on a low power system. (If you can use the the 80w brick, it is fanless - so probably prety quiet too.)
-
- Posts: 2
- Joined: Thu Nov 29, 2007 6:25 am
Is this machine an Everex TC2502 from Wal-Mart? If so, whats the noise level like from a few feet away?HedgeHocker wrote:More insight on a VIA processor based system... I own a C7-D with onboard VGA, two WD RE2 hdd's in mirror array, 2Gb Kingston 533Mhz, and a Plexwriter DVD-RW. My total system power draw acording to my Kill-A-Watt is 35-38w. I run this system 24/7 for torrents.
-
- Posts: 34
- Joined: Sat Apr 23, 2005 4:33 am
- Location: Sweden
- Contact:
I have a hard time believing that the energy consumption during manufacturing is greater than the energy consumption of an always on pc with at least 3 year of running.
For an average pc we can assume 100W for the core unit (no screen or printer etc.)that will ad up as 876kWh/year for three years that would be roughly 2,5MWh.
Is there any study that shows the energy consumption during manufacturing?
Edit:
If you assume that they can get really cheep electricity at the manufacturing sites its still a huge cost. If they can pay 2 cents per kWh 2.5MWh would amount to 50$ for electricity alone since you can by new MB for that price it seems unrealistic that the power consumption during production would be on a level with the power consumption of the end user.
For an average pc we can assume 100W for the core unit (no screen or printer etc.)that will ad up as 876kWh/year for three years that would be roughly 2,5MWh.
Is there any study that shows the energy consumption during manufacturing?
Edit:
If you assume that they can get really cheep electricity at the manufacturing sites its still a huge cost. If they can pay 2 cents per kWh 2.5MWh would amount to 50$ for electricity alone since you can by new MB for that price it seems unrealistic that the power consumption during production would be on a level with the power consumption of the end user.
First - a caveat, I have not read up on the details of this (but people like MikeC have.)I have a hard time believing that the energy consumption during manufacturing is greater than the energy consumption of an always on pc with at least 3 year of running.
Postings in the green computing forum and on ecopcreview give the impression that energy involved in manufacture computers is fairly significant compared to amount during use. e.g.
http://www.ecopcreview.com/LCA_and_ECPR?page=0%2C1But 96% of a refrigerator's lifetime fossil fuel consumption occurs during its use phase. For a computer, the situation is reversed: 25% occurs during use, while 75% occurs during production. So while its most important for appliances like a fridge to reduce energy use over the use phase, for computers, the best green strategy is to reduce energy in production and increase its useful life.
This is not to suggest that efforts to reduce a computer's electricity consumption by eco-minded users is in vain. But such efforts are not enough, from an ecological perspective. Buying products that require less energy to make, organizing to press manufacturers to become greener in the production, or simply extending the life of what you might consider a computer that's long in the tooth are all worthy options to consider, and they may be much more important that getting an extra 10W drop in AC consumption from a newer power supply or computer... especially when the ecological cost of manufacturing the new gear is considered. Reduce, Reuse and Recycle are still the 3Rs of eco-friendly behavior. In the case of computers, Reduce at the manufacturing level is paramount.
also see
viewtopic.php?t=40719
Of course more reading would be required to find out what vintage of computer is being considered here, lifetime, duty cycle, etc.
But even if the above was based on an 8hr day 5 days a week, going to on 24/7 would only increase the energy use perhaps 3-fold (less if it is idle much of the time). So the manufacturing and use energies would still be of similar magnitude.
A few thoughts about why that isn't so unreasonable.If you assume that they can get really cheep electricity at the manufacturing sites its still a huge cost. If they can pay 2 cents per kWh 2.5MWh would amount to 50$ for electricity alone since you can by new MB for that price it seems unrealistic that the power consumption during production would be on a level with the power consumption of the end user.
* A motherboard does not a computer make (aren't a lot of new whole computers you can get for $50.) The energy use mentioned is for the manufacture/transportation/... of the whole thing (i.e. a whole computer, so the $50 energy cost is part of an item which cost something like $500-2000+)
* Disposal/recycling costs are often not paid up front. (So if the energy involved there is included in the analysis, it may not factor into the purchase price.)
-
- Posts: 34
- Joined: Sat Apr 23, 2005 4:33 am
- Location: Sweden
- Contact:
I have some incite into electrical manufacturing industry but I have never heard any one taking about energy consumption as a cost adder.
I'm perfectly willing to agree that that the user case is very different from a refrigerator or a freezer but a life time analysis is extremely complex to do and often gives results that are very hard to interpret. And since MikeC doesn't give any references in the article that you quote its hard to validate his figures.
On the other hand I don't think this thread is the right "forum" for this discussion.
I will start a new thread in the Green computing forum when I have some spare time.
I'm perfectly willing to agree that that the user case is very different from a refrigerator or a freezer but a life time analysis is extremely complex to do and often gives results that are very hard to interpret. And since MikeC doesn't give any references in the article that you quote its hard to validate his figures.
On the other hand I don't think this thread is the right "forum" for this discussion.
I will start a new thread in the Green computing forum when I have some spare time.
-
- Site Admin
- Posts: 12285
- Joined: Sun Aug 11, 2002 3:26 pm
- Location: Vancouver, BC, Canada
- Contact:
The reference is right in there, just before the passage scdr quoted:prostetnic wrote:And since MikeC doesn't give any references in the article that you quote its hard to validate his figures.
That book is a kind of bible in its field.Among first published LCA studies of a desktop PC was Environmental Impacts in the Production of Personal Computers in 2003 by Eric Williams, then at UN University in Tokyo (and now at the Univ. of Arizona). It appeared as chapter three of the book Computers and the Environment, which Prof. Williams co-edited with Ruediger Kuehr. The study is long, complex and detailed, with a great deal of statistical analysis that hardly makes for compelling reading; however, its conclusions are compelling:
Here's a shorter piece by Eric Williams:
http://www.it-environment.org/publicati ... stract.pdf
-
- Site Admin
- Posts: 12285
- Joined: Sun Aug 11, 2002 3:26 pm
- Location: Vancouver, BC, Canada
- Contact:
MikeC wrote:The reference is right in there, just before the passage scdr quoted:prostetnic wrote:And since MikeC doesn't give any references in the article that you quote its hard to validate his figures.
That book is a kind of bible in its field.Among first published LCA studies of a desktop PC was Environmental Impacts in the Production of Personal Computers in 2003 by Eric Williams, then at UN University in Tokyo (and now at the Univ. of Arizona). It appeared as chapter three of the book Computers and the Environment, which Prof. Williams co-edited with Ruediger Kuehr. The study is long, complex and detailed, with a great deal of statistical analysis that hardly makes for compelling reading; however, its conclusions are compelling:
Here's a summary of a piece by Eric Williams:
http://www.it-environment.org/publicati ... stract.pdf
-
- Posts: 34
- Joined: Sat Apr 23, 2005 4:33 am
- Location: Sweden
- Contact:
-
- Posts: 34
- Joined: Sat Apr 23, 2005 4:33 am
- Location: Sweden
- Contact:
Ok I started a new thread for the discussion on energy consumption during manufacturing. viewtopic.php?p=382179#382179
Lets continue the discussion there.
Regarding my always on PC I'm currently thinking about getting a reused laptop if I can find a cheep one.
How hard is it to "silence" a laptop?
Can anyone suggest a laptop model/family etc that would work for my applications?
Lets continue the discussion there.
Regarding my always on PC I'm currently thinking about getting a reused laptop if I can find a cheep one.
How hard is it to "silence" a laptop?
Can anyone suggest a laptop model/family etc that would work for my applications?
My old server / desktop general purpose system had pentium 3 733MHz, 512MB ram, seagate 160GB PATA hdd, matrox g550 gpu.
It was the quietest machine I have ever not heard :p
Power usage was 42W idle and 60W max. If you want to go lower than that, you really get only a little diminishing returns vs the effort. Just save more on something else
It was the quietest machine I have ever not heard :p
Power usage was 42W idle and 60W max. If you want to go lower than that, you really get only a little diminishing returns vs the effort. Just save more on something else
Re: LOW Power PC for light aplications.
AMD S754 micro-ATX is the best desktop platform for low-power use. VIA CPUs stink from a performance/Watt perspective--their chipsets are generally pretty good though.prostetnic wrote:I need help finding a suitable motherboard and processor.
I’m thinking about going for a VIA EPIA micro-ATX but I’m open to any suggestion. Or should I go for a cheap laptop?
Any other sugestions for a system like this?
I am using a NEC Mobile Pro 900C, which is a fully solid-state, Windows CE micro-laptop (Handheld PC) that can do most of what you want to do with less than 5W of power--if you can stand using a 8" 640x240 passive matrix LCD. Not ideal for file serving (only has low-speed USB host port for external drives) and it probably can't burn DVDs. $100 or less on eBay.