Whoa, I had no idea Pineview was already available to reviewers. Been waiting for this, should really shake things up in the netbook market. Am reading the review now, will edit this later if I have anything to add.

edit: Power consumption is looking pretty good, can't wait to see some reviews of netbooks based on the N450. Am also kind of disappointed by lack of H.264/VC-1 decoding acceleration, but oh well.

Is the lack of video acceleration something definitive or it will be implemented sometime in the future? The new chipset surely offers some more power to do the task.
Also, maybe a useless question: how's the board BIOS? There's no mention of it at all in the review.

I kinda can't believe it - you are really praising the efficiency boost from the old to the new in your final thoughts? That's ridiculous. Nvidia with the ION showed what can be done - whereas INTEL only offered a crappy ass-old and outdated chipset along with its ATOMs.

The Pineview is not really an achievement - its would should've been out on the market from the beginning.

The only info. on that I have is the chipset leaves the option for a 3rd party HD decoder chip. The BIOS is just like the D945GCLF2: not worth mentioning.



The efficiency should be praised - have you read the reviews of the N450-powered Asus 1005PE? Battery life has improved a great deal thanks to this new design. We did point out that the IGP should be overhauled, but perhaps you feel our language was not strong enough. That's your perogative.


I don't see how that's possible considering Intel started consolidating their chipsets only recently.

Yeah I was just about to post that, apparently Intel has designed this to be paired with an optional Broadcom decoding chip. This is good news, hopefully this will be included, or at least available as an option, for most Pineview based systems. It sounds like the Dell Mini 10 based on the N450 will have the Broadcom chip as an optional component.

And I really don't know what people were expecting. The whole point of Pineview was to consolidate the CPU and GPU to reduce cost and power consumption. Atom has never been a high performance platform, were people expecting that to change with Pineview?

My point being, is just, that Intel had an old (very old) Chipset+IGP running together with the ATOM. The ATOM itself - no doubt - is efficient in its area of application.
So I'm just saying: It was about time, that INTEL released an up to date Chipset for the ATOM (2.0). You state "Boosting efficiency by that degree when the power draw is already low is very impressive" - that's just wrong because it is really not impressive to apply common modern technologie and the power draw wasn't that low compared to the competition.

But sure - without the old Chipset in the beginning, Netbooks probably would've been more expensive.

I know you got a preproduction model, but I wish they would have included HDMI out, or just drop VGA in favor of DVI-I so you could connect one of these to a TV with an adapter. This would be a good board to stick inside a Nintendo chassis and use as a retro-game rig/media extender (assuming x264 acceleration becomes available by vendor request)

I think an interesting comparison would be with the 945GSE-based board SPCR reveiwed here:

http://www.silentpcreview.com/Intel_D94 ... orex_T1610

The 945GSE uses less power -- but how much faster is the D510M0?

Perhaps you're thinking of this chip?

http://www.logicsupply.com/products/bcm970012

I don't think it'll be by much (at least thinking about the mobile N450). Well, percentage-wise, it's probably significant. In terms of practical application, the GMA3150 probably has very little effect.

Also note, the D510MO is the dual-core nettop version. We might very well see slightly better power consumption values for the N450.

God. I wish Intel would let them add DVI to this..... Ugh

A couple of the netbook manufacturers (notably the HP mini) have used a Broadcom crystalHD decoder board in a mini-PCIE socket to allow low performance Atom + 945GSE machines to decode HD video.

There's currently a fair bit of development underway to get myth and XBMC to use this hardware decoder, mainly to allow the use of an Apple TV box for front-end duties.

I note the motherboard tested has a mini PCI-E socket. It would be really interesting to see what sort of results the pineview based motherboard had when paired up with a HD decoder.

It might make a half-way decent myth/XBMC machine, running at very low power.

But you can't output full hd resolution in digital form. You only get LVDS (limited to 1366x768) and VGA (limited to 2048x1536, note N450 has a limit of 1400x1050 instead), no dvi/hdmi. It is possible though for board manufacturers (obviously not for netbooks) to use lvds->dvi adapter chip, but that doesn't get you more resolution.

Minor nitpick: The article refers to the name of the motherboard as D510M0 in some places and D510MO in others. May cause some people Googling for information to miss this article. I think the correct name ends with "MO" (as in Mount Olive), as per the Intel website

True, but look at the cost. This is a $75 board. I'm not sure what the MSRP for the D945GCLF2 is but I often see it for $80~85. Cheapest dual-core ION motherboard I've seen is $150 which is around double the price.


No, it's not an achievement, it's more a refinement. However, it should be remembered that Intel makes other processors and chipsets aside from the Atom and related. With the Atom, they're going for okay performance at low power consumption and cost. If you don't like the limitations, it's possible to build a considerably decent-performing Mini-ITX PC using a GF9300+Celeron E3200 combination.

I know that the D510MO is fanless, and has a total max power dissipation of only 33 Watts, but still I thought it was intended to be used in a case with at least a small amount of case fan induced airflow. Clearly, this review shows this board working fine in a morex T3500 case with no case fans, but the only reference to temperatures is that the external case and Atom heatsink got warm but "passed the touch test".

To the author: do you know the core temperatures under full load? And if so, were they any different with the case in desktop vs tower configuration? Also, what was the hard disk temperature? As someone interested in building fanless computers, even if I never build one with the D510MO, I still want to know the above as useful data points.

Thanks again for the review. silentpcreview is a great source of information.

These puppies are starting to get listed at halfway reasonable prices around here. But what cases/PSUs would be likely to be a good match? Do tell if I'm posting in the wrong forum.

It would be great if these boards could be run fanless with a low-power hard-drive and a modest amount of RAM.
We haven't seen temperatures but it looks like the Morex T3500 is up to the job, though I don't know if this pairing requires some kind of thermal pad like the one which was used in the Morex+D945GSEJT SPCR review. And if that case worked, why not another one with adequate airflow? The smaller Morex T3310 is cheap and available around here for instance but I don't know if it'd fit the D510MO, much less allow it to dissipate heat adequately. Does anyone have a clue? And is there another place than SPCR where one would be likely to find such information?
If nothing else, at least an M350+picoPSU from ebay should work, though that would only be marginally cheaper than a T3500 from ebay. It'd be a bit of a shame to pay more for that than for the board.

And if one were to drop the fanless requirement, would there any cheap but quiet solutions? Buying case, fan and PSU separately would likely end up being too expensive (maybe not for you North-Americans but we've got availability and/or shipping/customs to consider) and the cheap ITX cases/PSU combos I've seen look like they've got terrible airflow. I'm afraid their fans would be noisy.

this is the point. I have a D945GSEJT rig, that idles at 15w. So, slightly less than the new Pinewiew board.
I don't see here a consistent improvement in power consumption. We have to see if the N450 do it better.

And the Poulsbo boards consume even less. The point of the D510MO is that its cheap and fast (for an Atom) without consuming too much power.


mini-itx.com says the D510MO would fit in there so I'm tempted to buy this case for the sake of science.
Any tips for testing the heat dissipation capabilities of a fanless system, aside from waiting for the summer? I don't have a fancy thermometer or a greenhouse handy.

This is not a board you'd want to connect a display to except for troubleshooting, but it would be interesting to see how this or even more the little brother D410PT does as a server, e.g. running WHS. Wikipedia says the D410PT has 10W TDP while the D510MO has 13W TDP.
The D410PT may be better than the D945GSEJT for a WHS box. It also supports 64bit, which presumably WHSv2 will be.

Edit: Apparently D410PT does not have gigabit ethernet, what a useless design! D510MO has.

Just putting together a mobile system based on D510MO.

NO LVDS. The board has blank solder pads for it. C'mon, $1.00 worth of parts, and I could have used a spare laptop screen that's just sitting here on the shelf. All the websites show LVDS as "optional". Which means unavailable in Q < 1000. Bummer and Damnation.

My whole purpose is for survey data collection on battery power in an 11' open survey boat. Running off a 33AH lead acid gel cell, this machine will go all day FOR SURE including a 10-15W power budget for a small LCD. I am figuring it will come in under 4 amps @ 12'ish volts, probably averaging around 3A typical load, 2A idle with the LCD off, and jack shit in standby. I will cheat and add charging from the outboard, but it's not much extra, maybe 50% extra power in total, factoring in it's a small outboard that is off or idling much of the time. Still, a single battery should pull me through a typical 12H summer work day NO SWEAT.

My system: D510MO, PicoPSU M3, OCZ 30GB Solid2 SSD, 2GB DDR2@800, 802.11b/g wireless mini PCIe card, WinXP32.

I will probably be building into a large plastic box, allowing for complete and tidy storage + transport of the entire system including the battery and charger, all cords, monitor, KBD, etc.. Rain proof and bilge happy is my primary target, although this machine will be lovely silent in a hotel room for misc. evening internets duty (much better than my POS old laptop).

The box: MTM Case-Guard, orange, SPUD-7 18.5" x 13" x 15.25"

http://www.mtmcase-gard.com/products/camping/dry-boxes-spud6.html

These boxes are an amazing deal. Light, incredibly indestructibly strong, huge, immaculately well designed (they work very very well), and criminally cheap ($35). Pellican looses this race very badly on almost every count. No I'm not a shill, just a very happy user.

I have been using a single core Atom in the form of an Acer One netbook, and it's been quite capable in our covered boat, and even OK in the hotel for evening play / email. But it's cramped for screen and KBD, and fragile as far as things go. I will feel much better with an SSD and fully rainproof, and with a real monitor for spacious viewing.

My only other (silly) temptation is to build a "bagtop". I figure that since EVERY laptop depends on a much larger bag anyways, I can improve greatly on my user experience by simply building a mini-ITX based bag computer into a decent laptop bag. That way I can have more/bigger/better battery, and leave EVERYTHING all plugged in together all the time. No more fear of breaking off all those fragile connectors, USB sticks, etc.. No more power brick tangles, just pull the cord out of a side pocket and plug it in. Just zip open and use, that's the real target. The laptop is a limited, fragile, problematic, and inconvenient form factor to cram a computer into. A Panasonic Toughbook or an Itronix would be nice, but I can buy a good car for the $5000+ that they cost, and they still don't solve the problems I encounter with real world laptop usage in the field. I can probably even achieve a high degree of rain tolerance in a bagtop if I am careful with my build, and the convenience and open platform will be a geek's dream.

I'll let y'all know how the whole thing works, post some photos, etc.. Have parts, will build.

Sounds fun, crisspy!

btw, for your in-hotel times, consider this:
http://www.silentpcreview.com/LogicSupp ... eo_decoder

Thanks Mike, great little article. Sadly, there's a conflict with the 802.11 in that precious solo PCIe slot, and the single full size PCI will likely hold (ugh) quad serial ports (dinosaur era), so I can't make them trade places either. I guess I could go USB wireless, it's plenty cheap too.

And cheers mate, it's been way too long since I've been active enough around here. Sorry for being out of touch so bloody long too, but my hydrographic survey career has been chasing my ass all over BC for the last few years. No life, but the money is good for a change, and all in economic times when that is obviously far from the norm. I take it where I can get it, and my boss is a 110% best perfect guy to work with, so I won't complain about being employed in a job I honestly love. He buys lots of good toys (tools) too, and listens to me about those decisions. It's nearly a geek's heaven.

Meanwhile I am hitting the point of embarking on building some real field / boat computers. As per my post above, the norm is $5000+ Panasonic Toughbooks, that still have open leak-holes all over them, the second you actually need to actually use them, and open all the flimsy little hatches to actually plug something in. I think I can achieve a much more open and practical solution, even if it wouldn't look so good in 3MB .pdf marketing brochures to desk jockey idiot managers who authorize the big cheques. I guess it's another excellent niche market opportunity that will slide by me though, while I get on with getting surveys done.

For a hint on the theory vs. practicality divide, check out this bad boy:

http://www.logicsupply.com/products/systemgw_01

Beautiful piece of kit that, a mere $825. But really, a C7? and all those fancy pants connectors? Fine if you only ever wire it in once and NEVER touch it again. But I would have to spend around $1000 just to cable that beast properly, and end up right back in the land of cheap shit DB9, USB and RJ45. And if I did buy it, and build all the adapter cabling, the interconnections would still end up slopping around in a salty bilge. It's a practical fail, and lacks the real horsepower to run our software anyways. Nice (drool-worthy) case, wrong tactics.

So I'm heading to the land of cheap real field machines, with NORMAL cables, that live in rain and bilge (BC), and have enough power to do the work, and are made out of normal parts you can replace for normal prices. Overlaps quite a lot with silent computing, especially on points such as fanless and power efficient and DIY. And if the survey gig ever flops, I promise I'll start building & selling some of those quiet wooden cases as well as funky practical field machines. It's still a burning passion.

Alas, it's more than just that. All video outputs are limited in the resolution and none of them reaches full HD. Intel has purposely crippled the system. On a brighter note, the first generation Atoms are part of the embedded roadmap and will be available for quite a longe time (4~5years).

I had been looking at getting a D945GSEJT but now I am seriously considering this instead.

First, a couple of points about the review. The official name is D510MO, with a letter 'O' at the end instead of a zero. If you look at the pictures on Intel's web site you can see that the serial and parallel ports are just headers, so the lack rear connectors on your board is normal.

This board uses 5W more at idle than the D945GSEJT. That equals an extra 36p/month (at 10p/kwh) or £4.32 a year. Not bad considering the extra performance you get, plus the fact that you have two desktop DDR2 sockets.

I don't think the benchmarks do the dual-core Atom justice. In general desktop use dual-cores make things a lot snappier, and in server type roles the gains are even bigger. Truecrypt is almost twice as fast on 2 cores and throughput on the gigabit LAN for file serving will be a lot higher too, especially since both boards use a less-than-optimal Realtek LAN chip.

Replying to myself again...


I bought myself one of each. The board fits snugly in the case though the vents are perhaps not ideally placed. I like this case.
After laying it horizontally at 20.5C ambient, I let the hardware monitoring in the BIOS run with two RAM sticks, a keyboard and a display (nothing else). With the case open, the temperatures reported (processor and "remote") stabilized at 32.5C and 47C. With the case closed (but no drive installed), temperatures rose to 51C and 64C. Does anyone know where the "remote" sensor is located?
The top of the case got warm, but not uncomfortably so. SPCR's review said that the heatsink got warm to the ouch but not overly so. That's true but the chip next to it (the southbridge?) seemed to have gotten hot enough to burn, even with the case open. The AC adapter remained fairly cool to the touch, certainly compared with my laptop's brick.

I'll try more things with this box later and I'll supply SMART temperature readings. Do tell if there's any tests you'd like me to run.

And I'll finish by bumping my question:

Very interesting HFat. I wonder if a good quiet fan, say 80mm @ 5v, would stir things up enough just to keep temps in line. I mean, it doesn't matter how much or little heat sink parts have, no flow = heat builds at about 20W.

I have been thinking about strategy for wet marine environment, where it would obviously seem desirable to limit or prevent airflow on moisture grounds. Thing is, I will probably need to do something by way of active ventilation for the small space (a quiet fan), as your adventure with very small cases seems to suggest. And as for the moisture issue, it probably will be better too. According to Murphy (sorry to get religious here), we can count on any "sealed system" to leak, and end up stewing in the juices. So a 25W heater (the board itself) plus a fan might actually help to keep it drier than a closed case, out in the messy world. Not to mention that I don't see any truly practical way to close the thing up anyways, since it's hard if not impossible to truly seal the back panel of any mobo, no matter how many hours I might spend with crazy glue and caulking.

Cheers, and please do take my suggestion of one very super quiet fan as my choice of "test to run". Maybe a good glue-on heatsink for that SB chip too, if that makes any sense. I imagine a single fan wafting a hint of air past the CPU and SB (HS added), and venting out where it may, as probably being a practical minimum to keep these boards truly happy in the long run. At least in the realm of very small cases where large natural weather systems can't form on their own.

It's too early to tell given that I haven't looked at real-life temperatures without a fan yet. But, though I can feel a little heat coming from the box which I assume to be convection (could I feel such a small delta T of IR?), as much as I try I can't feel the air moving at all (my test is obviously set up in a place where there is no perceptible air movement). So I imagine the slowest of fans would change the picture.


I wouldn't jump to any conclusion based on my preliminary results. Keep in mind that this small case has large vents. It is difficult to see much of anything through the fine mesh that covers them so the components are protected from most solids but I assume liquids would easily penetrate the case.


Don't take this as a commitment on my part but... would you suggest specific parts?

Sorry no time to look myself, but NCIX and many others sell SB heatsinks, probably even kits that have HS + thermal epoxy. More later. Cheers.