Motherboards that have core2duo mobile processor support

coopers · Post by **coopers** » Sat Oct 13, 2007 12:52 am

Hi Folks,

New to the boards, but I've been reading for a bit, and I have some questions.

Firstly - I am interested in building a very quiet HTPC, and I've thought it might be a good idea to get a motherboard that supports a mobile processor. I started searching around for information on good motherboards for Merom intel core2duo's (socket M) and well... here I am.

So the questions:
- I see the thread at the top: "Recommended Motherboards: From a Silence Perspective 1.1" which is helpful, but all the motherboards in there are socket 775 motherboards, and there doesn't seem to be any discussion about what I am trying to do. So why is that? Is it a stupid idea? I would have thought this would be the place where people are running systems like I had in mind.
- I've had a search around in here for other people doing stuff with socket M full size motherboards... can't find a lot.
- I also see a lot of discussion about the AMD turion chips rather than the intel ones - how come? Are they more popular/better for some reason?

thanks

Coopers

jessekopelman · Post by **jessekopelman** » Sat Oct 13, 2007 2:20 am

Search the forums for MoDT (mobile on desktop). Right now there are hardly any boards that support dual core mobile processors. There are some good solutions if you can live with Pentium M.

The reason you see a lot of discussion about Turions is that the old ones used socket 754 and as such are compatible with most socket 754 motherboards. At this point, the availability of both the CPUs and the motherboards is not what it once was.

One note of interest is that pairing a mobile chip with a desktop board does not appear to lead to laptop level power consumption. My Pentium M laptop idles 12W, but I've yet to read of a MoDT Pentium M solution that does much better than 30W at idle. Meanwhile, plenty of people have been able to achieve 30W idles simply by undervolting AMD Athlons and Semprons. The current state of MoDT leads me to believe it is a bit of a fool's errand. At least in the case of Pentium M MoDT the prices aren't bad, though.

frank2003 · Post by **frank2003** » Sat Oct 13, 2007 2:58 am

I think the MoDT market is pretty much dead. I think what killed it are boards based on the inexpensive, highly underclockable/undervoltable AMD processors paired with feature rich chipsets like the AMD 690G and Nvidia 7050. Who would want to build a MoDT system when you can do it for one half the price using an AMD desktop processor?

hmsrolst · Post by **hmsrolst** » Sat Oct 13, 2007 6:03 am

There are some very good, reasonably priced MoDT solutions available as long as you stick to Yonah/Merom. There are very few Santa Rosa solutions currently.

I have two systems based on this AOpen board:

http://www.excaliberpc.com/Acer_AOpen_M ... 65365.html

and one based on this one:

http://www.excaliberpc.com/Acer_AOpen_M ... 69854.html

both for only a little over $100USD for the mobo.

In addition, a little careful shopping at ebay, and not aiming for the top of the CPU pile, will yield well-priced CPU's.

Note that both of the boards use standard socket 478 coolers, so systems based on these boards can be nearly silent. I agree that Pentium M solutions are also excellent, but Core Duo/2 Duo are just not that expensive.

smilingcrow · Post by **smilingcrow** » Sat Oct 13, 2007 2:15 pm

I wonder if the seeming failure of the MODT platform is based more on the lack of interest by OEMs in using and promoting these products in SFF systems, or even the lack of interest shown in them by mainstream PC users.

From my measurements the difference between a MODT mainstream system versus a similar C2D system is about 10W at idle with a good AM2 system being close to the MODT system; all being under-volted
At load though the S479 system extends its lead and the C2D system can even overtake the AM2 one. Using more extreme low power components (picoPSU, mobile drives, SSDs) will shrink the difference between the platforms further.

I think they all have their merits and I could choose any of them depending on the metrics that were most important when specifying a system.
Hopefully AMD will push their next generation mobile platform into the desktop space as it sounds very promising. I think it could be a good call for AMD as they have promoted the green credentials of their platforms for quite some time and there should be higher margins for this sector as well.

coopers · Post by **coopers** » Sun Oct 14, 2007 10:34 pm

Thanks folks, I will do a bit of research and report back.

jackylman · Post by **jackylman** » Sun Oct 14, 2007 10:50 pm

abit il-90mv
http://www.abit.com.tw/modt/il-90mv.htm

jessekopelman · Post by **jessekopelman** » Mon Oct 15, 2007 12:16 am

smilingcrow wrote: From my measurements the difference between a MODT mainstream system versus a similar C2D system is about 10W at idle with a good AM2 system being close to the MODT system; all being under-volted

So here's the real question, why can a laptop idle at well under 20W, but not any MoDT solutions? What tricks does the laptop motherboard employ that these MoDT motherboards don't? I think the big reason MoDT hasn't caught on is that the products offer no benefits that can't be achieved by any motherboard that supports undervolting paired with standard desktop CPUs. You'd think if Asus just ported one of their laptop motherboards unchanged onto a mini-ITX form factor they'd have a hit on their hands.

smilingcrow · Post by **smilingcrow** » Mon Oct 15, 2007 12:54 am

jessekopelman wrote:So here's the real question, why can a laptop idle at well under 20W, but not any MoDT solutions? What tricks does the laptop motherboard employ that these MoDT motherboards don't? I think the big reason MoDT hasn't caught on is that the products offer no benefits that can't be achieved by any motherboard that supports undervolting paired with standard desktop CPUs. You'd think if Asus just ported one of their laptop motherboards unchanged onto a mini-ITX form factor they'd have a hit on their hands.

I built a low power system with a MODT mobo with a Core Duo, PicoPSU and laptop drive and it consumed 22W at idle compared to 12W at idle for a Core Duo laptop with the screen disabled; a Core Solo reduced the desktop to 20W at idle. All systems were undervolted as much as possible.
Iâ€™m not sure where the extra 10W saving for the laptop comes from! Itâ€™s a large difference in percentage terms but not very significant in an absolute sense so I canâ€™t see many people paying much of a price premium for that.
The MODT boards donâ€™t allow under-volting in the BIOS which puts them at a disadvantage against AM2 systems. An AM2 single core system built to a similar spec that I outlined above should idle at 25W and probably lower which is good for cheap mainstream parts.

Itâ€™ll be interesting to see if Intelâ€™s Silverthorne platform migrates to the desktop as that looks like the beginning of a very lower x86 platform that should still have enough power for many situations; it supposedly will run as high as 1.7GHz and has 512kb L2 cache. The CPU and chipset will be very cheap for Intel to manufacture so may very well appear in mATX and mini-ITX format at low prices.

jojo4u · Post by **jojo4u** » Mon Oct 15, 2007 3:41 am

Well half of the way has already been done:
http://www.meisterkuehler.de/forum/ener ... erper.html
14,5 W in idle with hdd stopped (he's using a CF RAID now). Load 25 W with a CoreDuo U2500 ULV. Motherboard: iBase MB899, 1.8" Toshiba HDD, PicoPSU WI + LI SHIN 0225C1865 PSU.

http://www.meisterkuehler.de/forum/ener ... -idle.html
AMD X2 with Asus M2A-VM, PicoPSU-120: 21.5 Watt idle.
EDIT: He changed to a Asrock with 7050V and got 18 W now.

http://www.meisterkuehler.de/forum/ener ... rauch.html
VIA Epia M 10000 with 15.6 W idle. PicoPSU-120 used.

The first uses a ELV EM800 energy monitor, the 2nd and 3rdl use a Conrad Energy Check 3000.

smilingcrow · Post by **smilingcrow** » Mon Oct 15, 2007 4:05 am

jojo4u wrote:Well half of the way has already been done:
http://www.meisterkuehler.de/forum/ener ... erper.html
14,5 W in idle with hdd stopped (he's using a CF RAID now). Load 25 W with a CoreDuo U2500 ULV. Motherboard: iBase MB899, 1.8" Toshiba HDD, PicoPSU WI + LI SHIN 0225C1865 PSU.

But you arenâ€™t comparing like for like so it doesnâ€™t prove that much though the figures are still impressive.
A laptop with an U2500 and 1.8â€

jessekopelman · Post by **jessekopelman** » Mon Oct 15, 2007 12:17 pm

smilingcrow wrote: Iâ€™m not sure where the extra 10W saving for the laptop comes from! Itâ€™s a large difference in percentage terms but not very significant in an absolute sense so I canâ€™t see many people paying much of a price premium for that.

By that argument, why do MoDT at all? An undervolted desktop setup is certainly going to come within 10W of a MoDT setup and likely cost less. This is pretty much my point: Why do MoDT if it doesn't give you laptop-like performance?

smilingcrow · Post by **smilingcrow** » Mon Oct 15, 2007 1:26 pm

jessekopelman wrote:
smilingcrow wrote: Iâ€™m not sure where the extra 10W saving for the laptop comes from! Itâ€™s a large difference in percentage terms but not very significant in an absolute sense so I canâ€™t see many people paying much of a price premium for that.
By that argument, why do MoDT at all? An undervolted desktop setup is certainly going to come within 10W of a MoDT setup and likely cost less. This is pretty much my point: Why do MoDT if it doesn't give you laptop-like performance?

I see it as just another option that has a particular set of strengths and weaknesses.
A good AM2 system will get very close to a MODT system at idle but at load it just canâ€™t get close. A C2D desktop isnâ€™t that far away but offers more performance. AM2 is a cheap off the shelf platform. MODT can be very cheap from eBay. C2D is certainly more expensive but offers the performance. I feel that they all have their merits.

If you want the lowest power consumption at idle and load and require up to roughly 2.2GHz then MODT stands out.
You can even include the Q6600 G0 as a low power processor if you require that level of performance. With a laptop drive and undervolting you can just about get away with a PicoPSU and get an outstanding performance per watt for such a high level of performance.
This whole area is about niches and MODT fills one.

smilingcrow · Post by **smilingcrow** » Tue Oct 16, 2007 3:14 am

I forgot to mention that Intel is releasing a quad core mobile chip in H208 so thatâ€™s another niche that they are catering for. It has a TDP of ~45W and a FSB of 1066 so not compatible with current socket P chipsets.

GnatGoSplat · Post by **GnatGoSplat** » Tue Oct 16, 2007 4:48 am

I had a special need where I replaced the motherboard in a sort of embedded PC (media server from a now defunct company) and its custom, proprietary PSU is only 166W. As a result, power consumption under load was actually more of a concern than at idle.

jamesavery22 · Post by **jamesavery22** » Tue Oct 16, 2007 12:29 pm

Surprised that no one has mentioned the Asus N4L-VM DH.

There were only a hand-full of 945GM based mobos produced when the chipsets were first available. Most of said boards played on the whole ViiV thing. I still don't even know what the hell that thing is...
Anyways, mobile on desktop boards will continue to come out. C2D is a very slick platform,(desktop & mobile) but there is still enough drive in those niche avenues for ultra slim desktop computers to keep mobile desktop boards coming.

I don't think there have been any new mobile desktop mobos specifically because there hasn't been any new mobile chipsets. I believe the GM965 was released right around the same time as the 945GM. And there are a very small number of desktop boards that use it. Think an ITX one?
Regardless, until there is a new mobile chipset don't expect any new mobile on desktop boards to come out.

mattthemuppet · Post by **mattthemuppet** » Tue Oct 16, 2007 10:50 pm

some of the difference between similarly specced MODT and laptop systems is probably due to using SODIMMS and having only the bare essentials on the laptop mobo (2 SATA at most, 2 PCIe, 2 mem slots), though a few MODT boards do use SODIMMS (I think).

jessekopelman · Post by **jessekopelman** » Wed Oct 17, 2007 1:12 am

mattthemuppet wrote:some of the difference between similarly specced MODT and laptop systems is probably due to using SODIMMS and having only the bare essentials on the laptop mobo (2 SATA at most, 2 PCIe, 2 mem slots), though a few MODT boards do use SODIMMS (I think).

The SODIMMS aren't going to save you more than 1W/stick and likely only a fraction of that, so that can't be more than a small part of it. As for the other things, I've thought that might be the case as well, but why would these things draw power when not in use? I'm telling you, if you take a 945GM desktop system and configure it exactly the same as a 945GM based laptop (same CPU, same capacity and number of RAM sticks, same HDD) the desktop system is going to draw 10W more power. I think it's got to be the BIOS. I think there are certain Centrino features that need proper BIOS support. Whatever the case, I don't understand why someone, especially Asus who make laptops, can't just port a laptop motherboard exactly on to a mini-ITX form factor.

smilingcrow · Post by **smilingcrow** » Wed Oct 17, 2007 4:51 am

The Centrino laptops I tested consumed less than 1W on standby which is a useful saving compared to a desktop system; none of the ones I tested consumed less than 4W.

GnatGoSplat · Post by **GnatGoSplat** » Wed Oct 17, 2007 1:26 pm

I don't think SODIMMs are of any benefit whatsoever. I had some SODIMMs and regular DDR2 DIMMs that used the exact same chips. I prefer SODIMMs zero insertion force and their compact design (which I think should have been a standard for both desktop and laptop), but that's about the only benefit to them.

jaganath · Post by **jaganath** » Wed Oct 17, 2007 1:36 pm

a lot of power savings in laptops are probably from more efficient VRMs and suchlike.

jessekopelman · Post by **jessekopelman** » Thu Oct 18, 2007 1:15 am

jaganath wrote:a lot of power savings in laptops are probably from more efficient VRMs and suchlike.

I don't see VRMs as a good explanation for things like the vastly different (in a relative sense) standby draws. Again, signs point to a better implementation of sleep states in BIOS.

Back to my main point. Why can't a company like Asus use exactly the same parts and BIOS on it's MoDT board that it uses on its laptop boards? Even if this added $20 to the price, I don't think that would dissuade any customers. Meanwhile, better thermal and power draw performance would land a bunch of new customers who would otherwise go the undervolting route. The current state of MoDT is a half-assed effort that is not really targeted at the value conscious consumer. Not surprising considering it is pretty much a subset of the HTPC market, which is largely targeted at people with more money than sense. PC component makers are like most mature industries in that they'd rather differentiate products through marketing, rather than engineering. It's sad that heatsinks seem to get more product development thought than most components outside the microprocessors.

smilingcrow · Post by **smilingcrow** » Thu Oct 18, 2007 2:33 am

jessekopelman wrote:Back to my main point. Why can't a company like Asus use exactly the same parts and BIOS on it's MoDT board that it uses on its laptop boards? Even if this added $20 to the price, I don't think that would dissuade any customers. Meanwhile, better thermal and power draw performance would land a bunch of new customers who would otherwise go the undervolting route. The current state of MoDT is a half-assed effort that is not really targeted at the value conscious consumer.

Iâ€™m not convinced that saving the extra 10W is going to do much for MODT platform sales.
It would be good to see Asus produce an MODT board in the way you suggested but that still leaves the thorny issue of the power supply. For me this is more of a stumbling block as although the PicoPSU is a great product you still have the hassle of then sourcing a silent power brick with a known high efficiency. If there was an off the shelf ultra efficient power supply rated for 120W+ DC that could easily be incorporated into an ATX case I feel that would be an important step forward.
Manufacturers seem more interested in targeting the barebones SFF sector with MODT which is usually expensive, proprietary and has a limited upgrade path.

If enough people email manufacturers asking for certain products they might eventually produce them.

jessekopelman · Post by **jessekopelman** » Fri Oct 19, 2007 1:36 am

smilingcrow wrote:Iâ€™m not convinced that saving the extra 10W is going to do much for MODT platform sales.
It would be good to see Asus produce an MODT board in the way you suggested but that still leaves the thorny issue of the power supply. For me this is more of a stumbling block as although the PicoPSU is a great product you still have the hassle of then sourcing a silent power brick with a known high efficiency. If there was an off the shelf ultra efficient power supply rated for 120W+ DC that could easily be incorporated into an ATX case I feel that would be an important step forward.
Manufacturers seem more interested in targeting the barebones SFF sector with MODT which is usually expensive, proprietary and has a limited upgrade path.

If enough people email manufacturers asking for certain products they might eventually produce them.

I'm confused by this. Isn't the whole point of MoDT to achieve minimum possible draw? If I don't care about that, why am I going through the hassle/expense of using mobile CPUs? My point is that today's MoDT solutions are hardly better than undervolting a desktop CPU on a normal desktop MB. Shouldn't a MoDT solution be closer to a laptop level power draw than a tweaked desktop system? Also, what's with the 120+W PSU? My all-desktop-parts PC doesn't even draw 120W AC! If you are going MoDT, your goal should be well under 80W. There are plenty of good and inexpensive 80W and under AC/DC bricks. Going back to the Asus example, they could just give you the same ones they use with their laptops. Anyway, if you need 120+W and you want to use an ATX case, why not just get one of the several ATX form factor PSUs that are 80+% efficient at that draw? Maybe you want even higher efficiency, but such things really have very little to do with MoDT, which is by definition about using mobile spec'd components in a desktop application, not just more efficient desktop parts. Personally, I'd also rather just have more efficient desktop parts, but that is not the topic at hand.

jaganath · Post by **jaganath** » Fri Oct 19, 2007 3:50 am

Isn't the whole point of MoDT to achieve minimum possible draw? If I don't care about that, why am I going through the hassle/expense of using mobile CPUs?

why indeed. as you yourself point out:

today's MoDT solutions are hardly better than undervolting a desktop CPU on a normal desktop MB.

the bottom line is, if you want laptop-like power consumption, why not just buy a laptop? of course it would be much cheaper to use standard desktop parts and just undervolt the heck out of them, and the difference between that and the laptop system would only be a handful of watts at best, but if you're looking for absolute lowest power consumption then that's what you've got to do.

as for why Asus doesn't ship laptop boards as MoDT boards:
1) they won't fit properly into (m)ATX cases; the mounting points will be all wrong, same is probably true for mITX cases (which leaves open the question of why you would buy a MoDT board and then put it in a mITX case when that's basically what a laptop is)

2) can get higher margins selling those boards inside laptops than to the MoDT market (see point 3)

3) the MoDT market is already fairly tiny as it is; who needs such incredibly low power consumption and small form factor? only carPCs and scientific applications where they have to log data from experiments and suchlike, and this market is already well-served by the VIA EPIA offerings.

OK? can we stop flogging this dead horse now?

jessekopelman · Post by **jessekopelman** » Fri Oct 19, 2007 3:15 pm

jaganath wrote:the bottom line is, if you want laptop-like power consumption, why not just buy a laptop?

I can think of only one reason, but it is a reason more and more on peoples' minds -- because you want to build a HTPC that looks like a DVD player. You can probably get most of the functionality out of a laptop (support for 1080p might be a problem), but that would generally involve unsightly things like docking stations, external HDD, and USB TV tuners. You can make a great HTPC using a case like the Antec 2480 or Fusion, but those things are gigantic. How are you gonna pair that with your wall mounted flat screen? Of course that begs the question of what to do with your A/V receiver which is likely even bigger, if it's high end. Yes, the whole thing gets quite silly after a while and it becomes clear that a better solution may actually involve long cables and hidden components and have nothing to do with MoDT at all . . .

EDIT: I will note that if you need super small for your HTPC, MoDT is probably the way to go regardless of energy efficiency. Since the mobile parts have a lot more thermal tolerance, they are a much better choice for a case that has little to no room for fans and large heatsinks. Still, it seems like a ripoff that you don't get the energy efficiency too.

jaganath wrote:as for why Asus doesn't ship laptop boards as MoDT boards:

That wasn't my question. My question is, why doesn't Asus use exactly the same components and BIOS on their mini-ITX boards as they do on their laptop boards? If they are going to bother, why not do it right? Because they don't really care -- yes, I agree that is the likely answer.

jaganath wrote:OK? can we stop flogging this dead horse now?

I doubt it. Within the month there will be a new thread created on this topic.