Mini-review of Asus N4L-VM DH

[jamesavery22]

Sorry for the thread dig-up but this is one of the few N4L-VM threads that are worth anything and I wanted to make sure it was kept accurate

Merom consumes more than Yonah. Not the other way around. 2ghz C2D mobile's thermal envelope is 34w, 2ghz CD mobil is 31w.

Merom has a better performance/watt ratio. It consumes more, just outperforms much more...


Anyone else that stumbles across this thread should know there are a lot of people having problems with 2 x 1GB and using a PCI-E card, which includes well over half the people that use this board. Having two 1GB dimms while using a PCI-E graphics card has been reported to cause crashes. Asus is still putting out new bios' for this board so hopefully it will be fixed sooner rather than later, or never... Some people feel Asus is ignoring the issue.

I plan to just use one 1GB stick.

If anyone has found better solutions than the N4L-VM please chime in.

[jaganath]

Merom consumes more than Yonah. Not the other way around. 2ghz C2D mobile's thermal envelope is 34w, 2ghz CD mobil is 31w.

Remember TDP numbers are values in a range, not exact stipulations of how much a chip will use, Speedstep implementation is also very important, merom has lower Vcore minimum value than Yonah (1.0375V vs 1.1625V).

[smilingcrow]

That’s the minimum voltage when the processor is in the Highest Frequency Mode, so isn’t as much of an issue as it seems. The minimum voltage at idle (Lowest Frequency Mode) is 0.75V versus 0.7625V in favour of Core 2 Duo, which is negligible really.
Actually, I think that these values can be over-ridden by RMClock within limits; I ran a Core Duo at 1.66GHz/1.008V and 2.13GHz/1.15V, both of which are below Intel’s official lower limit of 1.1625V. Unless the software reporting the VCore was incorrect, which is always a possibility. It’ll be interesting to do some tests with the chips to see how they perform in the real-world.

[jamesavery22]

Yes TDP is just a limit, not a measurement. Majority of times Intel will slap the same TDP for a chunk of processors. Namely all P4 90nm chips from 2.4ghz to 3.2ghz have a TDP of 84w or 89w randomly. And we know a 3.2 sucks more power than a 2.4.
They are still good guides though. Actual consumption won't be too far from those numbers.

I have no idea what you guys are trying to get at about the Vcore other than being able to undervolt them Sounds like you are inferring Volts = Watts I'm sure thats not the case though. I've never seen anyone do any undervolting comparisons at all, let alone with some power consumption numbers.

Regardless, there are plenty of laptop reviews showing two equally configured laptops, one with a CD and the other with an equally clocked C2D and the CD having longer battery life. I havent seen any "true" tests in which the CPUs were just switched, but close enough. Its in line with the TDP #s and b/c of the big architecture difference it makes sense. Merom/Conroe is more efficient,64bit, and has more power saving features. Hence at full load merom is working much harder and consuming more energy. And at idle merom turns off more components and consumes less...

[Forumlurker]

higher volt = more Watt

http://forums.silentpcreview.com/viewto ... dervolting

you can see that higher volt gives mutch higher power consumtion...

[jamesavery22]

higher volt = more Watt

http://forums.silentpcreview.com/viewto ... dervolting

you can see that higher volt gives mutch higher power consumtion...

lol on the same processor yes... Thats like saying because you are giving 100v to a 100ohm resistor it is consuming more power than a 10ohm resistor getting 50v...

You are comparing two completely different loads. Ohm's law isn't exactly a hard concept come on

[jaganath]

The CMOS power equation is as follows (to a first order approximation):

P = 1/2 C V^2 f

where P is the power in Watts, C is the load capacitance in Farads, V is the supply voltage in Volts, and f is the clock frequency in Hertz. You can see from this equation that power consumption (and thus heat) is proportional to the square of the supply voltage, hence reductions in Vcore have a much greater effect on power consumption than reductions in operating frequency. Undervolting has been one of the key SPCR techniques for reducing power consumption ever since the the site was established.

[jamesavery22]

The CMOS power equation is as follows (to a first order approximation):

P = 1/2 C V^2 f

where P is the power in Watts, C is the load capacitance in Farads, V is the supply voltage in Volts, and f is the clock frequency in Hertz. You can see from this equation that power consumption (and thus heat) is proportional to the square of the supply voltage, hence reductions in Vcore have a much greater effect on power consumption than reductions in operating frequency. Undervolting has been one of the key SPCR techniques for reducing power consumption ever since the the site was established.

Yes I know. I am not disputing that... All I am saying is you cannot determine which core will consume more power by comparing their Vcore. Merom and Yonah are completely different cores. It doesn't matter that an equally clocked Yonah has a higher vcore. It still consumes less power because it is a different core, a lesser core if you will No one is saying lowering the vcore on any given CPU doesn't lower its power consumption. Really didn't mean to turn this into any form of discussion, thought it would be straight forward.

Isn't that the equation for the heat dissipation of a capacitor for each time its switched? Was under the impression thats just part of the equation for estimating heat dissipation in an IC... Maybe the rest of the equation doesn't have much impact? :shrugs:

[smilingcrow]

Merom and Yonah are completely different cores.

I wouldn’t go so far as to say that; I see Merom as being a heavily updated Yonah and built on possible exactly the same manufacturing process. So there are still strong connections between the two chips.

[jamesavery22]

Merom and Yonah are completely different cores.

I wouldn’t go so far as to say that; I see Merom as being a heavily updated Yonah and built on possible exactly the same manufacturing process. So there are still strong connections between the two chips.

Yes, true. I did exaggerate that. Merom has pretty solid roots from Yonah. They are definitely different though. Longer pipe, wider execution path. Merom/Conroe is more complex and there are plenty of articles that go over it.

Really didn't want to derail this thread... I'm still waiting to pull the trigger on a N4L-VM

[jaganath]

Isn't that the equation for the heat dissipation of a capacitor for each time its switched? Was under the impression thats just part of the equation for estimating heat dissipation in an IC... Maybe the rest of the equation doesn't have much impact? :shrugs:

Like I said, it's a first order approximation, the full equation looks like this:

P=1/2CV^2fN+QVfN+IV

N is signal switching coefficient, Q is charge due to through-type current, and I is leakage current. The last two terms are normally pretty negligible compared to dynamic switching power, although leakage current has been increasing in recent years (cf. Prescott).

I'm still waiting to pull the trigger on a N4L-VM

Is it not available where you are? The N4L-VM has been out in the UK for a couple of months.

[jamesavery22]

Kool, thanks for the EE info.

Yeah its available. I just started looking at 945 boards a few days ago so Im still reading on what available, whats good, whats not etc. Still reading on the 2 x 1GB dimm problems people are having on this board

Running out of time though so I'll probably just get this and try it out for myself. See a handfull of 945GT boards go up on ebay but not many. I've only seen 2-3 boards that I can actually buy that meet my reqs(PCI-E, onboard sound w/ 5.1 coax out, 2 PCI slots). The non-standard mounting is no big deal, was already going to hack up a heatsink or manufacture a new one.

Anyone else here running 945GM matx boards?

[jamesavery22]

This seems like a simple question but I cant find the answer... Does the N4L-VM support mobile C2D's? Their bios log shows an update to "support new CPU." And I know a lot of Dell laptops that used Core Duos just got BIOS updates to support Core 2 Duos.

Anyone?

edit:

Nice didnt know Asus had their own forum.

To answer my own question yes it does support it:

http://vip.asus.com/forum/view.aspx?id= ... uage=en-us

[smilingcrow]

This seems like a simple question but I cant find the answer... Does the N4L-VM support mobile C2D's?

It has better CPU support than most S479 motherboards as it also supports Celeron M 4xx series chips (Yonah based 65nm), which are very cheap in comparison to Core Solo CPUs. The Gigabyte S479 board doesn’t list Celeron M support; I’m gonna gave it a go anyway on the off chance.

Added: The Gigabyte board does support the Celeron M 420. Despite the lack of Speedstep support it hardly makes any difference to power consumption; 2W more at idle than a T2400. The difference between idle and full load is only 5W! This must be a low power bargain if you only need a single core.
The irony is that this chip has a TDP of 27W, when the whole system is consuming 44W at full load. 44W AC equates to ~31W DC with a Seasonic S12 430W; which puts this even more in perspective.

[kamina]

Anyone know if the N4L supports the T2300E? I'm thinking of building a very low voltage server for the house, and that combination seems tempting.

It's not on the official support list, that much I checked.

edit: actually the T2300E is suddenly not that tempting, there was an error in the article I read.

[kklein]

I just built a new system with one of these MB's and haven't experienced any of the problems other posters have mentioned.

I installed a Core 2 Duo T7200, 2GB of Kingston RAM (from the QVL list), and a 7100GS PCI Express card. Flashing the BIOS to support the C2D was no problem -- just put the BIOS on a floppy and press Alt-F2 during boot. I also haven't had any stability problems with the 2GB of RAM. Windows Vista has been running like a champ.

I now have a nearly silent system with excellent performance. I would highly recommend this MB to anyone.

[xaffeine]

My N4L runs great as part of an essentially inaudible system in an Antec P150 case. I have a Core Duo t2400 and the stock CPUFan.

It's been running almost a year now. In the early months, there were quite a few crashes. After about 3 BIOS upgrades, things have been very stable, quiet, and fast.

[Delta_42]

I'm running the Asus N4L-VM with a C2D T7200 1Gb of Ram and a 7600GT all placed lovingly into an Antec Fusion case. All runs nice and sweet although I'm thinking that the stock HSF is responsible for the fairly minimal noise coming from the PC. I'm looking at getting a Thermaltake HR-05 passive NB cooler to replace the CPU cooler (if it will fit in the case).

Apart from the joy of having to upgrade the bios to get the board to recognise the CPU this rig is making for a fantastic MPC, and whilst I have a dedicated rig for serious gaming there is something special about kicking back on the sofa with a PS2 gamepad playing GTA:San Andreas on the big TV, only now I'm doing it at 1280x768!

All I'm waiting for now is LG to release their HD-DVD/Blu-Ray dual format drive

[bean1975]

I have a T2300 in this board and the cooler is the venerable Zalman 3100. I have front and back fans in an Antec 3300 but no actual CPU fan. I just checked uptime, up 11 days and it was software that caused me to reboot not CPU heat.

You need the clip from the original cooler and trim it a little to fit the Zalman but it does work and it does fit if barely. The system has two 100G SATA HDDs stuck to the case with Scotch 111P Precut 1" Foam Mounting Squares, and two such squares has been applied on the cooler to separate the copper from the aluminium of the case -- otherwise it would touch it -- just to be on the safe side.

[Delta_42]

Scratch the HR-05, due to concerns whether it will fit my case, I'm now looking at getting a Noctua NC-U6. Will post pics once installed.

[jamesavery22]

hehe

Updates.

my N4L-VM project

[frostedflakes]

Wow pretty nice. What are temps like? Is that a 7900GT for the graphics card?

[jamesavery22]

thanks.

Idle temps

mobo - 30c, cpu 31c (core1 27c, core2 26c), GPU 49c. Thats completely idle with a 23-24c ambient. Load, CPU gets up to 50c, mobo stays below 37c. No clue on GPU. I dont have any games or any benching programs to load it. Its a 7900GS.

[Delta_42]

The Noctua cooler is finally in place. No whizzy little heatsink fan any more I'm now cooling my T7200 with a passive heatsink and the airflow from the two 120mm fans in the side of the case.

I'll get some pics before long.

Idle temps 23, load temp 55. difficult to tell the ambient temperature as I live in a draughty old house, there is probably a current of freezing cold air from outside blowing over my PC!

[audiojunk]

The Noctua cooler is finally in place. No whizzy little heatsink fan any more I'm now cooling my T7200 with a passive heatsink and the airflow from the two 120mm fans in the side of the case.

I'll get some pics before long.

Idle temps 23, load temp 55. difficult to tell the ambient temperature as I live in a draughty old house, there is probably a current of freezing cold air from outside blowing over my PC!

Stil waiting for pictures

I have this mobo too with a T2600,idle temps 32,Stressed(Orthos) around 43 with stockcooler,Q-fan diasbled,no powersheme and at 175fsb,180 gives a error after 6 minutes

[Delta_42]

Stil waiting for pictures

I have this mobo too with a T2600,idle temps 32,Stressed(Orthos) around 43 with stockcooler,Q-fan diasbled,no powersheme and at 175fsb,180 gives a error after 6 minutes

Ive done some pics of the Noctua in place, although it will be tomorrow until I can get them off the camera as I left the cable at work

I've got the back door of my house open at the moment, so my house is very cool... the T7200 is idling at 14 degrees!!

[Chocolinx]

The Noctua cooler is finally in place. No whizzy little heatsink fan any more I'm now cooling my T7200 with a passive heatsink and the airflow from the two 120mm fans in the side of the case.

I'll get some pics before long.

Idle temps 23, load temp 55. difficult to tell the ambient temperature as I live in a draughty old house, there is probably a current of freezing cold air from outside blowing over my PC!

Stil waiting for pictures

I have this mobo too with a T2600,idle temps 32,Stressed(Orthos) around 43 with stockcooler,Q-fan diasbled,no powersheme and at 175fsb,180 gives a error after 6 minutes

Have you tried overclocking the PCI-E Frequency a bit? Like to around 115? It should let you overclock a bit more, I've noticed this a lot on VM boards with integrated graphics. My system with P5LD2-VM, couldn't do more than 280 with the PCI-E on AUTO but I bumped it up to 115, then it let my FSB go up to 325!

Also I have a question for you, I plan on getting this board but I want to be sure about everything. Are you running the stock cooling without the fan? Maybe fanless? I recently grabbed this Sony Case (gutted it out) and it'll fit a HR-05. But I want to know if it's worth it to get one? I plan on sticking only one 120MM fan at 5V (Yate-Loon) and building some cheap ducts with hard plastic bags and bristleboard (like I did with my Centurion 5).

It's very unfortunate that you can't undervolt on this board. Unless someone over here knows of some mod?

[audiojunk]

The bios of the Asus N4L-VM DH is very limited,you can only adjust the fsb frequency and the ddr2 voltage
PCIe frequency is not available.
Undervoltage goes automatic with EIST,my T2600 clocks down to 1Ghz@0,95V and i'm running Internet Explorer and Mixmeister Fusion with it not bad for a 1Ghz processor.

[Benito]

For those of us still looking for a heatsink that will fit on Asus' brilliant mounting hooks on the N4L-VM DH:

I hope I finally found one: Thermaltake Pipe 101 rev.2 CL-P0006 (just rolls off the tongue, doesn't it?)

http://www.thermaltake.com/product/cool ... -p0006.asp

It's half a kilo of pure copper, the dimensions (67x68.5x47mm) are very close to those of the stock heatsink, and from the pictures it looks like the asus clip might just fit perfectly. If not, there is still the Amd xp clip to toy with. Only drawback is the spacing of the fins, which seems quite small.

As soon as I find one, I'll post the results.

[Delta_42]

This thread is a blast from the past, and I STILL haven't uploaded the pics of this mounted on my N4L-VM DH.

Its still going strong, I used the bracket off the stock HSF just looped over the Noctua cooler. This system is my HTPC and rarely gets turned off.

Obviously this sort of setup would only be any good in a case like the Fusion with good airflow close to the CPU.