pentium-m adapter for socket-478

Cooling Processors quietly

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee

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

Post by sgtpokey » Sat Mar 26, 2005 8:47 pm

Well as far as P-M desktop analysis goes, I prefer to look at functional criteria according to need. I already know I have two "desktop" computers that perform 2 different functions:

1. Game Computer
2. HTPC/File Server

So when I judge the P-M, I try to judge for those 2 different uses:

For #1, My priorites are : 1) Game performance 2) power draw 3) silence 4) price. So, looking solely at game benchmarks a P-M performs well. Additionally, if I used the EIST software legitreviews highlighted, I can even overclock dynamically, so that I can overclock when playing games, and underclock/undervolt when not. It's nice to know I've got this option. For a Game computer, I'd think about a P-M 1.86 or 2ghz machine: overvolted and overclocked when playing games, undervolted and underclocked when not. I may upgrade this summer from an Athlon XP, so the P-M may very well be my next gaming platform.

For #2, My priorites rearrange to 1) silence 2) power draw 3) performance 4) price. From this point of view, the P-M is a great architecture for the first two. For HTPC/file server performance: I currently use a mobile barton@6x200 = 1.2ghz for my HTPC so I know that any P-M or Celeron M will give adequate performance. I can also undervolt and underclock to really maximize power draw and silence. I do not know, however what the delta would be between my current setup and an M-based solution. The Celeron M may turn out to have better performance, better thermals, and a better noise floor, but I'm in no hurry to find out and I'm perfectly happy with my current setup.

Edward Ng
SPCR Reviewer
Posts: 2696
Joined: Thu Dec 11, 2003 9:53 pm
Location: Scarsdale, NY
Contact:

Post by Edward Ng » Sun Mar 27, 2005 8:08 am

B Bennion wrote:I'm hiding my HD's in the cat where the noise floor won't be noticed-- then the effort to go passive with the PM will pay off.... :?
Garfield?

Mats
Posts: 3044
Joined: Fri Dec 26, 2003 6:54 am
Location: Sweden

Post by Mats » Sun Mar 27, 2005 8:47 am

Edward Ng wrote:
B Bennion wrote:I'm hiding my HD's in the cat where the noise floor won't be noticed-- then the effort to go passive with the PM will pay off.... :?
Garfield?
:lol:

tay
Friend of SPCR
Posts: 793
Joined: Sat Dec 06, 2003 5:56 pm
Location: Boston, MA
Contact:

Post by tay » Thu Mar 31, 2005 4:22 am

GamePC has a look at it.

Mats
Posts: 3044
Joined: Fri Dec 26, 2003 6:54 am
Location: Sweden

Post by Mats » Thu Mar 31, 2005 7:06 am

tay wrote:GamePC has a look at it.
They're wrong when comparing 755 and 770.
The 7x0 is C0 revision and stepping 8,
the 7x5 is B1 revision and stepping 6,
the 130 nm CPU's got B1 revision and stepping 5.
As a side note, Celeron M is only available with the older B1 revision.

This time he PM looks better in performance:

The 4000+ wins over the 770 with 9 - 6
The 770 wins over the 3800+ with 8 - 7
The 770 wins over the 3500+ with 10 - 5
The 770 wins over the 3200+ with 12 - 3

Please don't just stare at the overclocked results, I did that myself until I realized that it just makes a big hype out of it. Since neither P4 nor A64 are overclocked it's very unfair. We all know that both CPU's are very good overclockers these days, too bad we didn't see a fair comparisation (again!). They didn't max out the FSB though, that would probably make PM look even better if they did. And yes, PM is still most low power!!!

Almost every time someone reviews a PM they uses the fastest CPU available and overclock it. When testing P4 or A64, the less expensive CPU's are used most of the times, which overclock quite good anyway.
What about PM 730 or 740, do they overclock good?

...oh yeah, GamePC is still a computer store....
Last edited by Mats on Thu Mar 31, 2005 7:49 am, edited 4 times in total.

tay
Friend of SPCR
Posts: 793
Joined: Sat Dec 06, 2003 5:56 pm
Location: Boston, MA
Contact:

Post by tay » Thu Mar 31, 2005 7:27 am

Excellent points Mats

PositiveSpin
Posts: 137
Joined: Sat Jul 24, 2004 5:03 am

Hiding hard drives in the cat?!?!?

Post by PositiveSpin » Thu Mar 31, 2005 9:12 am

B Bennion wrote:I'm hiding my HD's in the cat where the noise floor won't be noticed-- then the effort to go passive with the PM will pay off.... :?
Here's an interesting new technique with which I was previously unfamiliar - cat-mounted hard drives.

I went to the trouble of fitting my new low-noise machine with a mirrored pair of notebook drives (BTW: recommend this idea), when all I had to do was mount them in the cat.

I have four cats - can I RAID 5 them? Would that qualify as Cat 5?



I have several P4C800-E motherboards - I will be very happy to see them Pentium-M capable :-D

Mats
Posts: 3044
Joined: Fri Dec 26, 2003 6:54 am
Location: Sweden

Re: Hiding hard drives in the cat?!?!?

Post by Mats » Thu Mar 31, 2005 9:55 am

PositiveSpin wrote:I have several P4C800-E motherboards - I will be very happy to see them Pentium-M capable :-D
From Legit Reviews forums:
Well guys the P4C800E bios with voltage support has been released now.
The BIOS number is 1021 and is what I'll be using on the overclocking article.

tay
Friend of SPCR
Posts: 793
Joined: Sat Dec 06, 2003 5:56 pm
Location: Boston, MA
Contact:

Re: Hiding hard drives in the cat?!?!?

Post by tay » Thu Mar 31, 2005 11:20 am

PositiveSpin wrote:I have four cats - can I RAID 5 them? Would that qualify as Cat 5?
:shock: :D

Talz
Posts: 223
Joined: Mon Jul 07, 2003 9:29 pm

Post by Talz » Thu Mar 31, 2005 12:13 pm

I agree that to much shouldn't be made out of comparing overclocked to non oc'd chips, but it does merit some comparison as the P-M overclocked is easier to cool than an A64 or P4 at stock speeds and volts.

Mats
Posts: 3044
Joined: Fri Dec 26, 2003 6:54 am
Location: Sweden

Post by Mats » Thu Mar 31, 2005 12:39 pm

Talz wrote:I agree that to much shouldn't be made out of comparing overclocked to non oc'd chips, but it does merit some comparison as the P-M overclocked is easier to cool than an A64 or P4 at stock speeds and volts.
Sure, but the benchmarks tells us only about performance, not low power, and because of that it's still a very unfair review. People should be able to see what they could do with an overclocked A64 or P4 as well, as a comparisation. What about an overclocked $150 3000+ for instance?

Mats
Posts: 3044
Joined: Fri Dec 26, 2003 6:54 am
Location: Sweden

Post by Mats » Fri Apr 01, 2005 5:35 pm

The adapter at Legit Reviews have stopped working, and they have no clue why. Now we all know that the CPU and the mobo can handle overclocking, but we don't know anything about what the adapter can manage...

Shining Arcanine
Friend of SPCR
Posts: 502
Joined: Sat Oct 23, 2004 2:02 pm

Post by Shining Arcanine » Fri Apr 01, 2005 5:46 pm

Mats wrote:
Talz wrote:I agree that to much shouldn't be made out of comparing overclocked to non oc'd chips, but it does merit some comparison as the P-M overclocked is easier to cool than an A64 or P4 at stock speeds and volts.
Sure, but the benchmarks tells us only about performance, not low power, and because of that it's still a very unfair review. People should be able to see what they could do with an overclocked A64 or P4 as well, as a comparisation. What about an overclocked $150 3000+ for instance?
Regardless of what they can do they won't be able to do it while using less power than they do now (and without using less power it is virtually impossible to passively cool them) so it is irrevelent as far as silent computing is concerned.

Mats
Posts: 3044
Joined: Fri Dec 26, 2003 6:54 am
Location: Sweden

Post by Mats » Fri Apr 01, 2005 7:05 pm

Shining Arcanine wrote: Regardless of what they can do they won't be able to do it while using less power than they do now (and without using less power it is virtually impossible to passively cool them) so it is irrevelent as far as silent computing is concerned.
Not sure about that, we don't know what A64E is capable of yet, though some threads at Xtremesystems surely looks good.
They are only a couple of weeks away from showing up, you can preorder them even here in Sweden right now.
Passive cooling is still not the only way to get a really low noise computer.

Shining Arcanine
Friend of SPCR
Posts: 502
Joined: Sat Oct 23, 2004 2:02 pm

Post by Shining Arcanine » Fri Apr 01, 2005 7:27 pm

Mats wrote:Not sure about that, we don't know what A64E is capable of yet, though some threads at Xtremesystems surely looks good.
They are only a couple of weeks away from showing up, you can preorder them even here in Sweden right now.
I was referring to the processors avaliable now but if the new core can be passively cooled and overclocked simultaneously, then I see no reason not to compare its overclocked performance to an overclocked Pentium-M as in the end, they can be both passively cooled.
Mats wrote:Passive cooling is still not the only way to get a really low noise computer.
True but I didn't say it was. I said that it was irrevelent as far as silent computing is concerned and silent computing isn't low noise computing, it is no noise computing (according to the definition of silent).

Tibors
Patron of SPCR
Posts: 2674
Joined: Sun Jul 04, 2004 6:07 am
Location: Houten, The Netherlands, Europe

Post by Tibors » Fri Apr 01, 2005 7:38 pm

Maybe I missed something, but I had not noticed this thread was about passive cooling :?

There have been discussions of what is meant by silent at SPCR before. That definition in Webster does not come close to it. If you want to discuss this further with me please do so in one of those threads or start a new one. I'm sure I will find it.

Mats
Posts: 3044
Joined: Fri Dec 26, 2003 6:54 am
Location: Sweden

Post by Mats » Sat Apr 02, 2005 4:12 am

Tibors wrote:Maybe I missed something, but I had not noticed this thread was about passive cooling :?
Thanks Tibors!
I was into pure performance numbers when suddently someone told me that an overclocked A64 or P4 is not interesting at all? Yes we all know that PM is still most low power, but I don't think that's the reason why the reviewers overclock it (because you can do it and still be quite low power that is).

Shining Arcanine: Well the Asus CT-479 are not available either (just like A64E). For me there's no reason to cool a CPU passively rather than with a Zalman 7000 at min speed (not min Fanmate speed), I can't hear any of them inside a case so why bother? :wink:

Shining Arcanine
Friend of SPCR
Posts: 502
Joined: Sat Oct 23, 2004 2:02 pm

Post by Shining Arcanine » Sat Apr 02, 2005 7:00 am

Mats wrote:
Tibors wrote:Maybe I missed something, but I had not noticed this thread was about passive cooling :?
Thanks Tibors!
I was into pure performance numbers when suddently someone told me that an overclocked A64 or P4 is not interesting at all? Yes we all know that PM is still most low power, but I don't think that's the reason why the reviewers overclock it (because you can do it and still be quite low power that is).
And you overlooked the ability of the Pentium-M to be passively cooled which is something that the Pentium 4 and Althon can't claim to be even at stock.
Mats wrote:Shining Arcanine: Well the Asus CT-479 are not available either (just like A64E). For me there's no reason to cool a CPU passively rather than with a Zalman 7000 at min speed (not min Fanmate speed), I can't hear any of them inside a case so why bother? :wink:
The processor is avaliable and that is all I care about at the moment. As for there being no reason to cool a processor passively, if you can't hear fans inside a case that is good for you but I can and they drive me crazy.

Mats
Posts: 3044
Joined: Fri Dec 26, 2003 6:54 am
Location: Sweden

Post by Mats » Sat Apr 02, 2005 7:16 am

Shining Arcanine wrote:And you overlooked the ability of the Pentium-M to be passively cooled which is something that the Pentium 4 and Althon can't claim to be even at stock.
No I didn't. I just said that the discussion was just from a performance point of view. We were talking about heavy overclocked PM (which can't be cooled passively anyway) compared to other CPU's running at stock speed.
The processor is avaliable and that is all I care about at the moment.
But you still can't build the system this thread is about since you can't get the adapter yet, so it's just as theoretial as the A64E for now. Pretty useless with a processor and no mobo, right? :wink:
As for there being no reason to cool a processor passively, if you can't hear fans inside a case that is good for you but I can and they drive me crazy.
Please read my post again, I was pointing at a specific fan, not fans in general.

Edwood
Posts: 460
Joined: Wed Aug 28, 2002 4:48 pm
Location: My Secret Laboratory

Post by Edwood » Wed Apr 06, 2005 9:21 pm

There's also the issue of getting or modding a Heatsink to fit. The stock cooler is definitely not passive.

Asus should really have mounted the power input and jumpers underneath that adapter board. Hmmmm. I suppose they could be desoldered and soldered underneath.........

-Ed

Grubert
Posts: 31
Joined: Mon Apr 04, 2005 6:17 am
Location: Switzerland

Post by Grubert » Wed Apr 13, 2005 12:36 am

Tested on hardware.fr (in French).

google translation of the conclusions:
The use of Pentium M on desktop PC, confined hitherto with some impassioned users, could become well extensive with the arrival of Ct-479 d' ASUSTeK. This adapter indeed makes it possible to pass in addition to the problems involved in the mother charts nativement Pentium M: of a design which can appear sometimes limited, they were in more relatively expensive and not very available.

Whereas it was necessary before to pay more than 230 € for such a mother chart, the Ct-479 adapter, which should be available as of the weekend, will be to approximately 50 €, for which it will be necessary to add a hundred euros for a compatible mother chart if you do not have of them already one. Because it is the second effect kiss cool: Ct-479 makes it possible to extend the possibilities of upgrade of your machine, which are limited enough for that which have Northwood with a "low" frequency being given to it quasi unavailability of Northwood Socket 478 and in front of of Prescott Socket 478 which tend all the same to heat and which are limited to 3.2 GHz in commercial version.

It will be regretted however that ASUSTeK did not go yet at the end of this idea, by extending to the maximum the list of the compatible mother charts to charts such as the P4C800 and P4P800 not "- E", P4B533, P4G8X. On another side, before think of that it will already be necessary to perfect the bios currently supported mother charts, some being bugés and others not offering yet all the functionalities until one could wait - ASUSTeK indicates to work there actively. In the same way, the ventirad could have been of better quality, with for example an insert coppers on the level of the base a less audible ventilator on the level of its engine.

Of course, one should not lose sight of the fact either that Pentium M is not the perfect processor, although having a performances ratio/unbeatable thermal dissipation. From a technical point of view, Pentium 4 will be for example more at ease if you use mainly your PC for the video encoding. From a tariff point of view, it is necessary all the same to count 240 € for P-m 740 to 1.73 GHz - which is certainly very overclockable - that is to say the same price as Athlon 64 3500+. Knowing that because of the passage in 90nm and presence of Calm Cool', processors AMD Athlon 64, although a little greedier, can behave very well within the framework of a not very audible configuration, that clearly gives to reflect.

Without making the ideal Desktop solution of it, Ct-479 d' ASUSTeK thus return the solution Pentium M more accessible and powerful, while being equipped with some defects of youth. One can only greet ASUSTeK for this intelligent initiative which reduces to nothing the interest expensive and not very available mother charts dedicated, while hoping that the updates of bios will follow quickly.

hmsrolst
*Lifetime Patron*
Posts: 426
Joined: Sun Aug 11, 2002 3:26 pm
Location: Arlington, VA USA

Post by hmsrolst » Thu Apr 14, 2005 3:56 am

CT-479 now available in U.S. at:

http://www.xpcgear.com/asusct479.html

and

atacom

Shipping is about half as much at the first one.

jamesavery22
Posts: 271
Joined: Thu Jun 24, 2004 3:19 pm

Post by jamesavery22 » Thu Apr 14, 2005 1:38 pm

Mats wrote:...
No I didn't. I just said that the discussion was just from a performance point of view. We were talking about heavy overclocked PM (which can't be cooled passively anyway) compared to other CPU's running at stock speed.

...
I agree with you. When comparing two procs in terms of performance you can't OC one and not OC the other. But this isnt exactly the forum forum for pure performance. Im pretty sure GamePC's point in OC'ing the P-M was just this:
Image

Forgive the size and no compression, no image editing on this box =\

If a P-M could run at 3ghz it still would beat the 3200+ in terms of power consumption.

Thats a big deal to a lot of us who want to make the quietest box possible while still having some good performance. Sure I could get a dual xeon box that would blow it out of the water in number crunching. But the only people that would care are those that dont care about noise. OCing it to that speed says "Hey you can have a semi-powerful box and still use some really quiet solutions."

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

Post by sgtpokey » Thu Apr 14, 2005 2:33 pm

That power consumption graph (including the overclocking) and the gaming benchmarks are why I'm considering the P-M if I upgrade my gaming computer.

The funny thing is, as far as games go I think overclocking has a ceiling for usefulness. Games aren't like encoding where faster automatically equals (noticeably) better.
---
In any case, I'm now thinking:
1) slightly overclocked P-M for gaming and
2) undervolted/underclocked Mobile A-64 for HTPC/File Server

It's a fun time to think of upgrades. (I wonder if I'm the only one who's planning on upgrades before dual-cores/Windows 64 become mainstream)
Last edited by sgtpokey on Thu Apr 14, 2005 5:53 pm, edited 1 time in total.

Schlotkins
Posts: 278
Joined: Thu Mar 27, 2003 5:30 am

Post by Schlotkins » Thu Apr 14, 2005 5:00 pm

Those O'C'd PM numbers are _sick_. I mean, that's with a 6800GT! I wonder how much that eats up under load.

For me, I'm on a 2.4c and a 9600 np. It sounds like if I get a 2.0 or something like that, I should be able to get a good O'C out of it at at least be close to 3.73/FX-55 power at the same or less power consumption than I have now.

The question for me at this point is what do I use for a motherboard? It seems like (since I don't have one of the Asus boards) that I would _probably_ be better going with one of the real PM boards as I can use some serious heatsinks/fans. It seems I'd still be able to O/C and change voltage.

Chris

Mats
Posts: 3044
Joined: Fri Dec 26, 2003 6:54 am
Location: Sweden

Post by Mats » Mon Apr 18, 2005 1:56 am


hmsrolst
*Lifetime Patron*
Posts: 426
Joined: Sun Aug 11, 2002 3:26 pm
Location: Arlington, VA USA

CT-479 Mini-Review

Post by hmsrolst » Thu Apr 21, 2005 12:40 pm

My CT-479 was delivered a couple of hours ago, so I thought it would be helpful to share my experience.

I had an unused P4P800-VM, and I bought a 1.7 P-M at ebay. I also picked up a 70-80mm adapter from SVC, figuring I could replace the ASUS fan. In fact, the ASUS fan does just bolt to the heat sink using sheet metal screws, and some of the screws that came with the SVC adapter worked okay (slightly larger screws, which I'm sure could easily be picked up at a hardware store, would give a little more re-assuring grip). I added an L1A to the adapter, installed everything, and away it went.

As the reviews have described the CPU is immediately recognized with no problem, and, of course, the L1A at low speed is adequate for cooling even on the not-very-impressive heat sink. According to MBM the CPU is about 3 degrees higher than the mobo censor (35 vs. 32 at idle). As an aside, I now have three desktop P-M systems, and it's impressive that running Prime95 on the other two only raises the temps about 5 degrees.

I haven't had time to experiment with any of the freeware utilities that can lower the voltage, but overclocking the VM with Clockgen is straightforward.

If you already own one of the ASUS boards, so far this looks like a really good deal, since the main drawback of it for me was the 70mm fan.

CharlieChan
Patron of SPCR
Posts: 198
Joined: Sun Jul 13, 2003 2:57 am
Location: East Anglia, UK

Re: CT-479 Mini-Review

Post by CharlieChan » Thu Apr 21, 2005 12:54 pm

hmsrolst wrote:I had an unused P4P800-VM, and I bought a 1.7 P-M at ebay.
Is it a 479 pin or 478 pin 1.7 P-M?

Thanks.

hmsrolst
*Lifetime Patron*
Posts: 426
Joined: Sun Aug 11, 2002 3:26 pm
Location: Arlington, VA USA

Re: CT-479 Mini-Review

Post by hmsrolst » Thu Apr 21, 2005 2:52 pm

CharlieChan wrote:
hmsrolst wrote:I had an unused P4P800-VM, and I bought a 1.7 P-M at ebay.
Is it a 479 pin or 478 pin 1.7 P-M?

Thanks.
AFAIK it's a socket 478.

Post Reply