AOpen i915GMm-HFS: 2nd Gen Pentium M desktop board

Want to talk about one of the articles in SPCR? Here's the forum for you.
Ralf Hutter
SPCR Reviewer
Posts: 8636
Joined: Sat Nov 23, 2002 6:33 am
Location: Sunny SoCal

Post by Ralf Hutter » Wed Jun 22, 2005 4:53 am

sgtpokey wrote:
Not at idle it's not. I'm not pulling those power draw numbers out of my ass. They come from a Seasonic Power Angel, that my test systems are plugged into 24/7/365. I can see a drop of 2-3W under idle when going from default speed/Vcore to 600MHz @ .700V. As the CPU ramps up from idle to any sort of load, that 2-3W "advantage" quickly disappears.
Ok, just so I understand: the total power system draw of your P-M system at default speed and settings is only 2-3 watts more than when it was undervolted and underclocked?

I believe what you say but the total system power draw wasn't my focus with those comments and I originally misunderstand what you meant.

Now I'm curious however: do you know how your cpu temps varied during such a test (600mhz @ .7 volts versus default speed/voltage) or was it similarly unimpressive?
Re-reread my reply to you and you'll answer your own question. Here:
Yesterday Ralfie wrote:I can see a drop of 2-3W under idle when going from default speed/Vcore to 600MHz @ .700V.
To rephrase it:

At idle, and default speed and Vcore power draw is right at 38W.

At idle, and 600MHz, .700V the power draw is about 35W.

Temps showed maybe 1°C to 1.5°C (temp reading was bouncing back and forth) cooler under minimum power/speed settings. Too small of a difference to be meaningful.

FWIW, this has been consistent across all three different Pentium M boards that I've used.

Ralf Hutter
SPCR Reviewer
Posts: 8636
Joined: Sat Nov 23, 2002 6:33 am
Location: Sunny SoCal

Post by Ralf Hutter » Wed Jun 22, 2005 5:00 am

Dominic wrote: and my assessment is that we here should not be overly put-off by the HSF. After all, most of us would change it, even were it identical to their good designs. :lol:
I'll reiterate what joemadeus stated in his reply to you:

You can't just change the stock heatsink because the mounting hole pattern is proprietary and nobody has been able to find an aftermarket heatsink that will bolt on w/o major modifications. In other words, there are no alternative heatsinks for this board.

Ralf Hutter
SPCR Reviewer
Posts: 8636
Joined: Sat Nov 23, 2002 6:33 am
Location: Sunny SoCal

Post by Ralf Hutter » Wed Jun 22, 2005 5:06 am

elec999 wrote:Two question how do you mount a different heatsink with this board.
As I've stated multiple times, in this thread alone, you don't. The mounting holes on this board are a non-standard pattern, so no aftermarket heatsink will just bolt right on. If you wanted to have a machine shop drill a new set of holes into an existing heatsink, you may be able to get something to fit, but that's an absolutely ridiculous (IMO) hoop to have to jump through to get your $300 motherboard to keep the CPU somewhat cool.
elec999 wrote:And second why is it so expensive.
Thanks
I dunno. Ask AOpen.

My guesses, 1) It's a very full-featured, cutting-edge board, and 2) they may not expect to sell millions of them so the cost-per-unit will be higher.

Dominic
Posts: 31
Joined: Fri Jun 17, 2005 8:43 am
Location: Cambridgeshire, United Kingdom

Post by Dominic » Wed Jun 22, 2005 5:31 am

Hmm, that seems to put aopen in a bit (an understatement if ever there was one) of a tight spot! For, most people here are interested in a CPU which dissipates the the least heat (well, not the least, rather which does so most efficiently), and as such, this board would hence fail entirely. I do, however, think some might overlook the heat thing, if they decide that the feature set is right for them.

I really like the holistic approach it seems this board takes: it's MATX, it has 2 x gigabit ethernet, HDTV output, etc. However, at the temperatures it runs, I'd say aopen should merely have swapped the plans/dimensions of the northbridge HS with those of the CPU HS - as was noted within the review, the northbridge one seems larger! :lol:

On another, more off-topic, note, I'm now seeing this all in 1600*1200 goodness, at 100hz - my new 22' CRT is sweet!

gavino
Posts: 1
Joined: Sat Jun 18, 2005 8:25 am

Post by gavino » Wed Jun 22, 2005 5:35 am

Great review Ralf.

I really wished they'd get rid of the following:
1. Onboard FDD headers
2. Onboard PATA connector
3. Onboard Parallel Port and Comm ports
4. Onboard PS/2 header

USB 2.0 and SATA removes all need in my mind for this kit. You can get USB IR dongles, USB port replicators / converters for legacy IO like Comm ports and parallel ports and USB adapters for any old-school keyboards. If you have a non-USB mouse give it to someone else and buy a USB one I say. Sampling Hz are better ;) Modern BIOSes can boot from USB-FDDs - and I mean how often do you need to use these nowadays?! I use my FDD for exactly one reason only - loading my Silicon Image drivers during a windows install. I generally flash BIOSes and firmware (DVD-RW and GPU) off a DOS-bootable USB key.

These 'improvements' (IMHO) would leave more room onboard for a decent heatsink solution.

As for PATA - MSI and Plextor are making S-ATAPI drives now, and as other posters stated you can get SATA- 2.5" HDDs.

I wish they'd make legacy-free motherboards for ppl like me! In about, say, 3 years I'm sure they will. (Apple now that you're going x86....please! :)

joemadeus
Posts: 32
Joined: Fri May 28, 2004 10:38 am

Post by joemadeus » Wed Jun 22, 2005 5:37 am

Ralf Hutter wrote:
elec999 wrote:Two question how do you mount a different heatsink with this board.
As I've stated multiple times, in this thread alone, you don't.
I was able to make the mods I needed with hand tools. Copper is soft and fairly easily drilled, though I now owe my girlfriend a new drill bit. To tell truth, I can see how I could have owed her a new drill, since if you go too fast the copper gets gummy and can make low-power equiptment sieze. Also, a drill press would have made things go a LOT easier.

The other thing to watch out for is the base thickness. I happened to find a copper sink with a base about 1/32" thicker than the AOpen one. Countersink the bores if you need to, to get it to the correct thickness. You risk damaging the CPU otherwise.

I totally agree that a $300 board a) came with such a lousy sink and b) has a nonstandard pattern. Really - why not just use a NB sink pattern, at least?


-j

sgtpokey
*Lifetime Patron*
Posts: 301
Joined: Sat Dec 14, 2002 11:29 pm
Location: Dublin, CA / Liverpool UK

Post by sgtpokey » Wed Jun 22, 2005 5:59 am

Temps showed maybe 1°C to 1.5°C (temp reading was bouncing back and forth) cooler under minimum power/speed settings. Too small of a difference to be meaningful.
ok thanks. Yes, I already understand that the power draw of the entire system didn't change much. I was more interested in any localized temp effect on the cpu chip itself.

FWIW, I'm still designing an ultra low-power, distributed setup, so I'm trying to get a handle on individual contributors of heat and am not so concerned about total power draw (it's a distributed system with the hard drives external and power brick external; the only heat generators inside the case would be the motherboard itself (including on board video), tv tuner card, optical drive, and the cpu).

Ralf Hutter
SPCR Reviewer
Posts: 8636
Joined: Sat Nov 23, 2002 6:33 am
Location: Sunny SoCal

Post by Ralf Hutter » Wed Jun 22, 2005 6:10 am

sgtpokey wrote: FWIW, I'm still designing an ultra low-power, distributed setup, so I'm trying to get a handle on individual contributors of heat and am not so concerned about total power draw (it's a distributed system with the hard drives external and power brick external; the only heat generators inside the case would be the motherboard itself (including on board video), tv tuner card, optical drive, and the cpu).
I don't know if this is entirely germane to your project, but I've been using my P-M system to do a lot of burning, and a lot of transferring files from CDr's to HDD. My Plextor DVD burner actually uses more power under load than the 2.0GHz CPU uses under full load. If I run CPUBurn, my total system power draw is around 55W. If I'm leaning on the DVDrom drive (and not loading the CPU at all), the system power draw goes up to 58-59W. So FWIW, unless YMMV with your particular optical drive, or usage patterns (boy, how's that for qualifiers? :) ) you can plan on your optical drive putting out more heat under load than your CPU. This situation will be exacerbated if you undervolt your P-M CPU. Dropping the Vcore from default (1.324V) to 1.100V drops my full load power consumption to 48-49W.

sgtpokey
*Lifetime Patron*
Posts: 301
Joined: Sat Dec 14, 2002 11:29 pm
Location: Dublin, CA / Liverpool UK

Post by sgtpokey » Wed Jun 22, 2005 6:26 am

Ralf,

That's pretty funny regarding optical drive power usage. I guess that's why on mini-itx forums they always caution you to use notebook-style dvd drives due to their lower power requirements. I do plan on using a spare notebook dvd-rom and have no need for a burner.

This aopen board would be great for these plans were it not for the crazy non-standard heatsink mounts.

MikeC
Site Admin
Posts: 12285
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Post by MikeC » Wed Jun 22, 2005 6:35 am

Funny how no one seems tro have noticed the replacement HSF AOpen is sending out that I mentioned yesterday. :?:

It looks at least twice as big as the original, which is hopeful. RH should have it today. My guess is that it may be or is already the new "standard issue" HS for this board. IE, AOpen IS responding to customer feedback (and SPCR review criticism) on this issue.

IsaacKuo
Posts: 1705
Joined: Fri Jan 23, 2004 7:50 am
Location: Baton Rouge, Louisiana

Post by IsaacKuo » Wed Jun 22, 2005 6:43 am

Ralf Hutter wrote:The mounting holes on this board are a non-standard pattern, so no aftermarket heatsink will just bolt right on.
Could an aftermarket VGA cooler fit--maybe a Zalman flower cooler? I know Zalman's heatpipe VGA coolers are very adjustable, but the flower cooler seems to have a fixed bracket with a couple nonadjustable (elongated) holes. Whether this fits or not would depend on the exact distance between the mounting holes.

Dominic
Posts: 31
Joined: Fri Jun 17, 2005 8:43 am
Location: Cambridgeshire, United Kingdom

Post by Dominic » Wed Jun 22, 2005 6:48 am

MikeC wrote:Funny how no one seems tro have noticed the replacement HSF AOpen is sending out that I mentioned yesterday. :?:

It looks at least twice as big as the original, which is hopeful. RH should have it today. My guess is that it may be or is already the new "standard issue" HS for this board. IE, AOpen IS responding to customer feedback (and SPCR review criticism) on this issue.
I must have missed it! :oops:
In that case, I am most pleasantly surprised, and for me, it further confirms what I've always thought about aopen and that in my opinion they're the best motherboard manafacturer! (I'd give up my abit kv7 for an aopen board anyday, after the troubles I've had!)

grandpa_boris
Posts: 255
Joined: Thu Jun 05, 2003 9:45 am
Location: CA

Post by grandpa_boris » Wed Jun 22, 2005 7:15 am

MikeC wrote:Funny how no one seems tro have noticed the replacement HSF AOpen is sending out that I mentioned yesterday. :?:
i, for one, noticed it. to me it seemed like a minor improvement. i am comparing this new aopen HS to something like scythe ninja or zalman 7x00, which can possibly run passively, and it doesn't seem to be in the same league.

from my perspective, the right solution is to provide for a standard HS mount holes, allowing for 3rd party HS installation. aopen had specifically chosen to go with a mounting option that is unique and incompatible with any other HS on the market. this could have been an accidental oversight on the part of the design team, or it could have been a deliberate marketing decision. the former suggests a lack of product competent management oversight, the latter suggests product management's nearsightedness and conceit. i suspect that in neither case it is possible to fix the situation without an expensive retooling effort to relayout the board.

MikeC
Site Admin
Posts: 12285
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Post by MikeC » Wed Jun 22, 2005 7:29 am

grandpa_boris wrote:i, for one, noticed it. to me it seemed like a minor improvement. i am comparing this new aopen HS to something like scythe ninja or zalman 7x00, which can possibly run passively, and it doesn't seem to be in the same league.
Those are not really relevant comparisons, imo, because of the size issue. Remember that this board is meant to be a small all-in-one solution. Also, fanless CPU cooling is not necessarily ideal. More and more, I think it's a detriment because of the potential for components overheating on the motherboard. This causes a broad range of issues, including reduced operational life and higher potential for early failure for the board, significantly reduced efficiency (and increased power consumption), and often, much higher CPU temps, which also affects the PSU sitting (usually) directly above it. All the old timers know by now how often I've reiterated that a very quiet fan with a wee bit of airflow is usually acoustically indistinguishable from no fan at all, and much much cooler.
.... i suspect that in neither case it is possible to fix the situation without an expensive retooling effort to relayout the board.
Precisely why something like this bigger HSF might be the only viable solution... though it is too bad they could not have gone to copper and make it possible to mount an 80mm fan.

Dominic
Posts: 31
Joined: Fri Jun 17, 2005 8:43 am
Location: Cambridgeshire, United Kingdom

Post by Dominic » Wed Jun 22, 2005 7:48 am

grandpa_boris wrote:to me it seemed like a minor improvement. i am comparing this new aopen HS to something like scythe ninja or zalman 7x00, which can possibly run passively, and it doesn't seem to be in the same league.
I personally feel that running anything passively is in nobody's interests. One really needs some form of airflow if one is to cool powerful semiconductor devices, even if only a case fan. Firstly, one must realise that computers were, and are, not designed to be run passively. They are specifically designed to be used in conjunction with some form of active cooling. Secondly, and perhaps more importantly, I'd make the case that although it might seem quieter when being run passive, one can make just as quiet (not in scientific terms, but at least audibly speaking) a machine by use of some well placed active cooling, and as a result maintain far lower system temperatures, and better stability to boot (no pun intended).

A valiant effort this is from aopen, as I see it. But as a passive system? Well, let's just say that I'm not convinced! :wink:

sgtpokey
*Lifetime Patron*
Posts: 301
Joined: Sat Dec 14, 2002 11:29 pm
Location: Dublin, CA / Liverpool UK

Post by sgtpokey » Wed Jun 22, 2005 8:28 am

sonofdbn wrote:
Does anyone know what temp the Pentium M runs at inside a laptop? I'm just wondering because on laptops there isn't a lot of space for cooling (although there is certainly heat dissipation through the casing) and apart from the occasional fan whine I don't hear much either.
To answer your question, a 2ghz Dothan quickly > 55c without an active fan. After that the fan kicks in on my laptop so who knows how much hotter it could have gotten.

grandpa_boris
Posts: 255
Joined: Thu Jun 05, 2003 9:45 am
Location: CA

Post by grandpa_boris » Wed Jun 22, 2005 8:32 am

Dominic wrote: I personally feel that running anything passively is in nobody's interests.
i have 2 general purpose passively cooled systems in my home office. if i count the dedicated devices (firewall, DSL modem, printer server), the count goes to 5. i very much prefer silent to quiet.
One really needs some form of airflow if one is to cool powerful semiconductor devices, even if only a case fan.
in cases where convection airflow isn't sufficient, some form of active air circulation is essential. that doesn't in any way mean that you should have a fan attached to every hot component in your system. it should be possible to design the airflow and the cooling surfaces in a way that will minimize the number of fans.
I'd make the case that although it might seem quieter when being run passive, one can make just as quiet (not in scientific terms, but at least audibly speaking) a machine by use of some well placed active cooling, and as a result maintain far lower system temperatures, and better stability to boot (no pun intended).
i think this is all highly subjective. i am very sensitive to mechanical noises, to the point that i use battery digital alarm clocks rather than AC ones because the 60Hz hum of a typical AC alarm clock's cheap power supply bothers me. anything i can do to lower the computer's noise level below ambient is worth a consideration for me. so a possibility of having a reasonably powerful system that can be operated passively or with the fewest number of fans is of interest to me.

grandpa_boris
Posts: 255
Joined: Thu Jun 05, 2003 9:45 am
Location: CA

Post by grandpa_boris » Wed Jun 22, 2005 8:49 am

MikeC wrote: Those are not really relevant comparisons, imo, because of the size issue. Remember that this board is meant to be a small all-in-one solution.
i realize that. however, my impression is that aopen could have provided for a more conventional HS mount form factor and still had a custom mount for their proprietary HS.
Also, fanless CPU cooling is not necessarily ideal. More and more, I think it's a detriment because of the potential for components overheating on the motherboard. This causes a broad range of issues, including reduced operational life and higher potential for early failure for the board, significantly reduced efficiency (and increased power consumption), and often, much higher CPU temps, which also affects the PSU sitting (usually) directly above it. All the old timers know by now how often I've reiterated that a very quiet fan with a wee bit of airflow is usually acoustically indistinguishable from no fan at all, and much much cooler.
in my view, having a forced air circulation through the case doesn't imply or require having a CPU cooled by a dedicated fan.

Dominic
Posts: 31
Joined: Fri Jun 17, 2005 8:43 am
Location: Cambridgeshire, United Kingdom

Post by Dominic » Wed Jun 22, 2005 9:51 am

grandpa_boris wrote: in my view, having a forced air circulation through the case doesn't imply or require having a CPU cooled by a dedicated fan.
I'd agree! Perhaps I misread you, but I thought that by "passive", you meant an entirely passive system - i.e., one that has no forced airflow.

However, I'd still stick by the idea that one really should have some dedicated form of active CPU cooling, even if just for the extra airflow over the northbridge heatsink that the said cooling provides. It seems far more efficient to concentrate the power of cooling solutions upon the areas which generate the most heat!

MikeC
Site Admin
Posts: 12285
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Post by MikeC » Wed Jun 22, 2005 9:54 am

grandpa_boris wrote:i realize that. however, my impression is that aopen could have provided for a more conventional HS mount form factor and still had a custom mount for their proprietary HS.
Agreed.
in my view, having a forced air circulation through the case doesn't imply or require having a CPU cooled by a dedicated fan.
Certainly not in all cases, and probably not in the context of the low power draw of the P-M, but there are clear benefits. Both Intel and AMD specifically refer to airflow blowing down on the CPU HS and creating airflow around the VRMs for normal operation of the motherboard.

hmsrolst
*Lifetime Patron*
Posts: 427
Joined: Sun Aug 11, 2002 3:26 pm
Location: Arlington, VA USA

Post by hmsrolst » Wed Jun 22, 2005 12:48 pm

Although this board should be a lot better with the new HSF, it still seems awfully expensive for the modest advantages of the 915 chipset over the 865. For me this review confirms the growing superiority of the ASUS CT-479 solution. I currently am running 3 P-M's on P4P800 variants, as well as 2 on AOPEN 855 boards.

ASUS is slowly producing more flexible BIOS's, the ASUS HS works very well with a 70-80mm adapter, and Zalman 7000 and 7700 HSF's seem to be easily modifiable to work with the CT-479.

My latest setup is a P-M 730 (1.6GHz, 533MHz FSB) on a P4P800SE. The latest BIOS allows under and overvolting vcore, along with FSB and multiplier adjustments. I just ran Prime95 for six hours with the 730 at 10 x 201 undervolted to 1.25v. (My chip doesn't seem to be a particularly good overclocker--some are getting to 2.4 with default voltage, whereas mine barely squeaks into Windows at 2.4 with 1.5v.).

The upshot is that for half the price of the AOPEN 915 you get dual channel RAM which can run at PC3200 or PC4000 (at 201 FSB and above there's a 5:4 ratio) speeds, along with temps under load in the mid-40's. Then again, I have no need for PCI-E.

wim
Posts: 777
Joined: Wed Apr 28, 2004 5:16 am
Location: canberra, australia

Post by wim » Thu Jun 23, 2005 1:15 am

i'm with gramps on all counts
(incidentally i also bought a battery operated digital clock because the old one's noise was getting on my nerves. it's comforting to know i'm not the only nutcase in the world!)
Dominic wrote:I personally feel that running anything passively is in nobody's interests. One really needs some form of airflow if one is to cool powerful semiconductor devices, even if only a case fan. Firstly, one must realise that computers were, and are, not designed to be run passively.
while i agree with most of the things you mention, i don't think you should speak for other people's interests. if someone wants the aesthetic value of no-fans, even if there is no audible improvement, then sure it's unscientific of them..but if it's what they want, then it's in their interests. if that's what they get their hearts set on, likely they won't be satisfied until they have fanless operation even if it ain't the rational engineering solution.

while it is true that computers are designed with active cooling assumed, this doesn't need to be true for future designs. i don't think you would argue that it's impossible to have semiconductor devices as "powerful" as todays computers (in the usable sense) cooled passively, with the correct design. would you...? it should just requires sufficient attention to efficiency and optimal cooling which has been missing in the past. heck, even if it were impossible, not bothering to try and even aim there is not going to help bring about a silent computing revolution..

Dominic
Posts: 31
Joined: Fri Jun 17, 2005 8:43 am
Location: Cambridgeshire, United Kingdom

Post by Dominic » Thu Jun 23, 2005 2:49 am

wim wrote:while i agree with most of the things you mention, i don't think you should speak for other people's interests. if someone wants the aesthetic value of no-fans, even if there is no audible improvement, then sure it's unscientific of them..but if it's what they want, then it's in their interests. if that's what they get their hearts set on, likely they won't be satisfied until they have fanless operation even if it ain't the rational engineering solution.

while it is true that computers are designed with active cooling assumed, this doesn't need to be true for future designs. i don't think you would argue that it's impossible to have semiconductor devices as "powerful" as todays computers (in the usable sense) cooled passively, with the correct design. would you...? it should just requires sufficient attention to efficiency and optimal cooling which has been missing in the past. heck, even if it were impossible, not bothering to try and even aim there is not going to help bring about a silent computing revolution..
Firstly: i agree utterly and wholeheartedly that choice is important! If one wants an entirely passive system, let one have it!

Secondly: I do not believe the future of silent computing to lie in passive cooling, rather in more efficient and less noisy active solutions. There is a massive difference between an entirely passive system, and one with even one case fan blowing the hot air away. Unless one could find a way to disperse heat from each of the components to the four corners of the room the pc is housed in, using heatpipes or the such, there is just too much heat concentrated in one place to make passive cooling viable - bar any advances in passive cooling technology - and that heat must somehow be dispersed over a wider area! Think of it: all the heat from your pc staying in your pc or, in the case of passive cooling, being blown away, and circulating throughout your house! The amount of air which the heat comes in contact with in the latter is exponentially larger and, when one thinks about it, is worth the price of having one silent fan bothering and nagging one at the back of one's mind.

However, I'll reiterate what I said earlier in this post: If one want passive, I would not stop one from having it!

Buzku
Posts: 16
Joined: Mon May 16, 2005 2:04 am
Location: Finland

Post by Buzku » Thu Jun 23, 2005 3:47 am

THG reviewed the same board, their temps were quite different:

"With our settings, the CPU fan was always off during normal Windows use. In fact, the processor never got any warmer than 43°C. Watching a DVD resulted in a system load of about 25%, but even this only increased the temperature by one degree."

Same HSF that was in your review.

Ralf Hutter
SPCR Reviewer
Posts: 8636
Joined: Sat Nov 23, 2002 6:33 am
Location: Sunny SoCal

Post by Ralf Hutter » Thu Jun 23, 2005 4:02 am

Buzku wrote:THG reviewed the same board, their temps were quite different:

"With our settings, the CPU fan was always off during normal Windows use. In fact, the processor never got any warmer than 43°C. Watching a DVD resulted in a system load of about 25%, but even this only increased the temperature by one degree."

Same HSF that was in your review.
I call bullshit, big time.

No one other than Tom's has got these sort of temp numbers with this
board. Read through this thread and visit the links provided for other reviews, as well as the many people posting on AOpen's own forum that are having high temp issues with this board.

Search SPCR for "toms" or "THG" and you'll quickly find that Toms is held in extremely low regard around here. Their testing and methodology is generally quite flawed. To make things even worse, they are just not capable of being objective when it comes to Intel products because their noses are buried so deeply up Intel's bunghole. Anything that Tom's says about an Intel product should be taken with a giant grain of salt, IMHO.

Dominic
Posts: 31
Joined: Fri Jun 17, 2005 8:43 am
Location: Cambridgeshire, United Kingdom

Post by Dominic » Thu Jun 23, 2005 4:36 am

Ralf Hutter wrote:Anything that Tom's says about an Intel product should be taken with a giant grain of salt, IMHO.
Can't we make that a half-grain? Lets not give them too much credit! (yes, I know that salt is a giant ionic structure and is hard to break up - but, in a special case like this, we could try ;) )

sgtpokey
*Lifetime Patron*
Posts: 301
Joined: Sat Dec 14, 2002 11:29 pm
Location: Dublin, CA / Liverpool UK

Post by sgtpokey » Thu Jun 23, 2005 5:14 am

... there is just too much heat concentrated in one place to make passive cooling viable - bar any advances in passive cooling technology -
Or the other way of achieving it is if you build a computer with lower overall thermals and not top-of-the-line processing power. That way you don't have as much concentrated heat in the first place. Such designs are here right now in the mobile and small form factor markets (Fanless Tablet PCs and mini-itx htpcs). Not disagreeing with the gist of your position, but the definition of high-performance computing is changing as computers get into more markets. (Or maybe I should say the computing market is segmenting into raw high powered cpus and thermally friendly cpus)

But in any case in the summer I'll be attempting to build a fanless htpc, so I'll be testing in practice some of the theories regarding the viability of fanless designs.

OOPs, I went off topic. Ok so this aopen board is out of the running due to the non-standard mounts since I'm leaning on using mcubed's BORG heatpipe cooler in this build (in conjuntion with mcubed's heatsink case)

Dominic
Posts: 31
Joined: Fri Jun 17, 2005 8:43 am
Location: Cambridgeshire, United Kingdom

Post by Dominic » Thu Jun 23, 2005 5:34 am

sgtpokey wrote:Or the other way of achieving it is if you build a computer with lower overall thermals and not top-of-the-line processing power. That way you don't have as much concentrated heat in the first place.
Precisely! One must tackle the issue of heat at the source, rather than trying to compensate for it via cooling. That was, effectively, what I was attempting to say when I said that modern semiconductors are designed for active cooling! ;)

wim
Posts: 777
Joined: Wed Apr 28, 2004 5:16 am
Location: canberra, australia

Post by wim » Fri Jun 24, 2005 1:15 am

partly OT..
Think of it: all the heat from your pc staying in your pc or, in the case of passive cooling, being blown away, and circulating throughout your house!
i think you meant to say active in this sentence, but ok - passive cooling doesn't necessarily have to be a closed system. this is just a drawback of standard computer cases, that it would tend to seal the air in a hot box like that. but with the correct design and/or geometry, natural convection could also be exploited to circulate airflow (granted it's a very weak effect). and where natural convection is sufficient, it is the correct choice - you wouldn't want active cooling on, say, LCD power brick (currently warming my toes) just because it works much, much better.
Dominic wrote:
sgtpokey wrote:Or the other way of achieving it is if you build a computer with lower overall thermals and not top-of-the-line processing power. That way you don't have as much concentrated heat in the first place.
Precisely! One must tackle the issue of heat at the source, rather than trying to compensate for it via cooling. That was, effectively, what I was attempting to say when I said that modern semiconductors are designed for active cooling! ;)
gynah..well actually i meant the same thing when i said
it should just require sufficient attention to efficiency...which has been missing in the past
we're beginning to see the start of this with P-M boards for desktops like this. actually i'm quite amazed at the thermal characteristics of dothan and winchester/venice cores compared to recent predecessors! if the trend extends to other chips and components we could realistically get passive cooled computers which consume drastically less power but are at least as "powerful" as today's. if they can be efficient enough to make passive cooling feasible, then passive is more desirable imo (less moving parts and less noise, more 'elegant'). MikeC's beef with passive cooling is that it's not really a good trade-off with current computer hardware. OK, fair. but it can still be a goal for future hardware, right? and if nobody is aiming to get passive cooled computers with today hardware then nobody working in industry is going to see a market for pushing efficiency to the limits (to try and design something to sell it to those guys and make their task easier. demand comes first, then supply)
Secondly: I do not believe the future of silent computing to lie in passive cooling, rather in more efficient and less noisy active solutions.
well maybe i'm more optimistic than you. i can see tiny passive cooled computers when i'm my parents age - ok not the near future, but maybe the distant future, i think (hope)!
maybe wishful thinking but then again, i mean the main limitations should be semiconductor design issues and i don't think we're close to hitting up against any theoretical boundaries yet, because superfluous heat output hasn't been seriously attacked until recently..

mco_chris
Posts: 58
Joined: Wed Dec 29, 2004 9:04 pm
Location: Orlando FL

Post by mco_chris » Mon Jun 27, 2005 7:18 am

I was really excited about this board when I first heard about it. I was going to buy it for my HTPC - the DVI out and Pentium M seemed just right for me.

The proprietary CPU heatsink mount is a deal-killer. I can't understand why they did that.

I'm interested in hearing about the new Aopen heatsink, but it's too little, too late.

I already own the first version of the Aopen Pentium M board and like it very much.

Hopefully the next Pentium M boards will allow us to use standard CPU heatsinks.

Post Reply