Page 3 of 5

Re: WMVHD Test

Posted: Mon Jun 27, 2005 7:57 pm
by vinylvision
It sure would be helpful to me if someone would answer Hyphe's questions. If the MB were to require a separate video card to drive a HD projector at 1080p resolution, the MB would not be able to my needs - heat and other issues aside. Thanks (This is a great forum)
Hyphe wrote:Great review. To bad about the HSF...
Could anybody tell me, or make a test if this motherboard, with the built in graphics, could play a WMVHD 1080p movie? I am planning to build a HTPC and I am not really sure what type of processor to buy for it to be able to cover playback of most formats. Would, for example Pentium M 745 be enough? :?:

Posted: Mon Jun 27, 2005 8:04 pm
by sthayashi
Umm... I thought the ability to decode a 1080p WMV-HD was CPU dependent, not GPU dependent.

Posted: Mon Jun 27, 2005 8:17 pm
by vinylvision
You could be right. I'll start another thread tomorrow rather than butting in on this one. I want to include some HW "requirements" info from the Joe Kane Productions website.
sthayashi wrote:Umm... I thought the ability to decode a 1080p WMV-HD was CPU dependent, not GPU dependent.

Posted: Tue Jun 28, 2005 3:29 am
by Mariner
A lot of the newer GPUs provide WMV-HD decode assistance which takes some of the load off the CPU. In fact, I'd say that if you're talking about 1080p, such support would be almost necessary as the amount of processing to be done could be too much for all but the fastest CPU to cope with.

I've no experience of using the integrated graphics in the 915 but I'd be surprised if it provided any assistance to WMV-HD decode.

Posted: Sun Jul 03, 2005 8:58 am
by Hyphe
The integrated graphics does not seem to be that impressive so I think I'll have to get a seperate GPU anyway...
MikeC wrote:Anyway, since the review was posted, I've been contacted by AOpen about a new, improved HSF for this board, which is being sent to both RH and myself -- RH will test its efficacy and I will measure / record the noise.
Did you recieve the heatsinks from Aopen? Will it be updated soon?

Posted: Mon Jul 04, 2005 5:13 am
by Ralf Hutter
Hyphe wrote:The integrated graphics does not seem to be that impressive so I think I'll have to get a seperate GPU anyway...
MikeC wrote:Anyway, since the review was posted, I've been contacted by AOpen about a new, improved HSF for this board, which is being sent to both RH and myself -- RH will test its efficacy and I will measure / record the noise.
Did you recieve the heatsinks from Aopen? Will it be updated soon?
I got mine, and will test it when I have the time and inclination.

I spent a lot of time (actually, "wasted a lot of time" might be more appropriate) working with the stupid original heatsink, and when I was done with the review I tore down the 915 setup and rebuilt my earlier 855GMe setup. At this point, I am not too inclined to rip it apart again, just to test this latest attempt. I'll do it, but don't hold your breath.

Posted: Mon Jul 04, 2005 6:53 am
by dragonii
Hi Ralph,

Just to tell you that I'm another interested guy on your future(?) testing of the HS. BTW, no pressure...

Just a bit frustating to be running on the 60°C on idle for a Pentium M 1.6 Mhz...

Regards

Posted: Mon Jul 04, 2005 7:13 am
by Hyphe
Hehe ok, I understand...

I got some answers to the question I had earlier, take a look at this thread on avs:
http://www.avsforum.com/avs-vb/showthread.php?t=555685

Posted: Wed Jul 13, 2005 4:50 am
by dragonii
Probably some hope for us owners of this MB:

http://www.ko-soku.co.jp/sales/cooler_775.htm

Posted: Thu Jul 14, 2005 2:17 am
by fucoboy
dragonii, re your link: nice list of HSs. Do you think that one of those would match?
thankx

Posted: Thu Jul 14, 2005 3:49 am
by dragonii
Look for the our motherboard name and you'll find the one that aparently matches. Now, if only I could order it... :?

Posted: Thu Jul 14, 2005 4:21 am
by Ralf Hutter
dragonii wrote:Look for the our motherboard name and you'll find the one that apparently matches. Now, if only I could order it... :?
Looks like all they're selling is an adapter that lets you mount an Alpha PAL6035 on that board.Since I can't read Japanese, I can't tell if they're selling the four year old Alpha along with the adapter.

Passive cooling for the pentium-m

Posted: Sun Jul 17, 2005 11:34 am
by Hyphe
Here's a great alternative to you guys who bought this motherboard but can't find a good cooling solution. It's made for their own case but could fit other cases/cooling solutions.

A passive heatpipe solution adapted for the 479 socket, they even have a picture showing it mounted on a i915Gmm-HFS... :shock:

http://www.t-balancer.com/catalog/produ ... 677a0ce595

Posted: Mon Jul 18, 2005 1:02 am
by dragonii
Thanks for the lin. Seems promising (no fan).

Who's willing to take the plunge and try out one ? I was looking into the Zalman Cu range as mentionned in the article as a desperate measure but I'm unsure if the Zalman will fit into my HTPC case (have the 5.25 bay a few cm above the CPU). Well, I'll probably order one of these and see how it goes...

Posted: Mon Jul 18, 2005 8:33 am
by chiccoweb
Regarding the cpu cooling I sort it out in this way:

Image

http://forums.silentpcreview.com/viewtopic.php?t=23742

Now I hope Microcool made a NorthPole with a new clip :wink:

Posted: Tue Jul 19, 2005 1:17 pm
by fucoboy
Mr. Hutter,
first of all I would like to thank you now for your great review. I really appreciate your profound approach. I have tried to reach similar results as you by undervolting my 760 CPU to 1.1V. But even that didn't prevent my CPU from idling at a temp of 60C and more. Under max load the temperature now seems to be stable at 70-75C with fan running at 100%. Maybe I need to reinstall the heatsink. BTW, where did you place those rubber shims, below the board or on top between board and heatsink? I have them on top but they do not seem to serve for anything.

Unfortunately I found out that undervolting my system seems to lead to a loss of stability. Running Prime95 Torture Test failed after a short period of time (<1 min.). Now having set to voltage back to its default value of 1.308V the torture test runs fine (since several ours by now without an error). Maybe this is also caused by my ultra-low-latency memory (2 GB GEIL 533MHz, 3-3-3-8), but I am not sure.

As many others I am still looking for an alternative cooler solution. The ones recently posted by chiccoweb and Hyphe are looking promisingly, particulary the passive one. Now I am looking forward to any testing results.

Thanks to all who are testing and trying out things are share their experiences...

Posted: Wed Jul 20, 2005 3:49 am
by Ralf Hutter
fucoboy wrote: BTW, where did you place those rubber shims, below the board or on top between board and heatsink? I have them on top but they do not seem to serve for anything.
I placed them between the top of the stiffening bracket and the underside of the mobo. I also tried leaving them out altogether, hoping for better contact between the HS and the CPU die, but that made no difference in my temps.
fucoboy wrote:Unfortunately I found out that undervolting my system seems to lead to a loss of stability. Running Prime95 Torture Test failed after a short period of time (<1 min.).
Undervolting too much can lead to the exact same type of instability as if you were overclocking too much without enough Vcore. It's just as important to check the stability of an undervolted system as it is to check an OC'ed system.

Every CPU is different, and just because mine will run stable at 1.1V doesn't mean yours will. I suggest you try again. but start out at 1.15V or something. If that errors out in Prime95, raise it to 1.175 and try again. Eventually you'll find a stable setting and will be able to run it there.

Posted: Wed Jul 20, 2005 4:34 am
by hmsrolst
Ralf Hutter wrote:
fucoboy wrote:Unfortunately I found out that undervolting my system seems to lead to a loss of stability. Running Prime95 Torture Test failed after a short period of time (<1 min.).
Undervolting too much can lead to the exact same type of instability as if you were overclocking too much without enough Vcore. It's just as important to check the stability of an undervolted system as it is to check an OC'ed system.

Every CPU is different, and just because mine will run stable at 1.1V doesn't mean yours will. I suggest you try again. but start out at 1.15V or something. If that errors out in Prime95, raise it to 1.175 and try again. Eventually you'll find a stable setting and will be able to run it there.
Just to confirm Ralf--I have a half dozen P-M's, and all will undervolt stably in the 1.1-1.2 range, but not all will do 1.1.

Posted: Mon Jul 25, 2005 7:33 am
by fucoboy
Thanks to you all. I really appreciate your help...

Posted: Sun Jul 31, 2005 11:01 am
by Hyphe
dragonii, did you try anything? Which solution did you order?

Posted: Sun Jul 31, 2005 10:37 pm
by dragonii
Hype,

Ordered one but they only come in the 15 of this month. I'll keep this thread updated when I finally receive it.

Posted: Tue Aug 02, 2005 3:47 am
by shleepy
First of all, great review, Ralf!

I have a couple questions. What's up with the i915Ga-HFS? Basically, it's an ATX version of the board? If so, does the HSF still mount non-standardly?

Secondly, would anyone be willing to mod a heatsink for me that would work well on this board? I've heard good things about modding pretty basic heatsinks, so perhaps a mid-high end one would be the one for me. I'd obviously pay you for your work and I hear it takes under an hour. I'm just a little concerned about trying to do that myself and do not even have the proper tools to drill through thick copper, for example.

Oh, and finallly - would a 7800GTX fit in here? Obviously, it's a little cramped and whatnot.

Anyway, if you're willing to help me by modding a heatsink, please reply. Thank you!

POSTSCRIPT: IMPROVED HSF

Posted: Sun Aug 07, 2005 8:09 am
by MikeC

Still seems hot

Posted: Wed Aug 10, 2005 6:42 am
by mc2wheels
61 degrees still seems really hot for the cpu under load. Not sure they made that much of an improvement. I'd rather see them redesign the mounting bracket to use after market copper heat sinks. This board was really intriguing, but a pentium-m chip & board is too expensive to run hotter than an athlon 64 venice combo with a decent heat sink. Just my 2 cents.

Re: Still seems hot

Posted: Wed Aug 10, 2005 7:21 am
by joemadeus
mc2wheels wrote:61 degrees still seems really hot for the cpu under load. Not sure they made that much of an improvement. I'd rather see them redesign the mounting bracket to use after market copper heat sinks. This board was really intriguing, but a pentium-m chip & board is too expensive to run hotter than an athlon 64 venice combo with a decent heat sink. Just my 2 cents.
I wonder if this is the sensor. I mounted a modified Thermaltake Volcano 10 (all copper) with an 80mm panaflow (medium airflow version) at 12v, blowing through a 80mm-60mm adapter, and I STILL get 55-57C temps when running Prime95. It's been anywhere between 25-35C in my house over the past few weeks, fwiw.


-j

Posted: Wed Aug 10, 2005 7:41 am
by Caudipteryx zoui
This board is still at least 100 dollars too expensive, even with dual gigabit LAN and DDR2 RAM (both of which I don't need). Combined with the also-pricey Pentium M, this setup gets an extreme beatdown when it comes to the price/quiet performance ratio of a C&Q AMD64 w/ mATX board (lacking only gigabit LAN and DDR2).

Re: Still seems hot

Posted: Thu Aug 11, 2005 7:44 am
by Ralf Hutter
joemadeus wrote:
mc2wheels wrote:61 degrees still seems really hot for the cpu under load. Not sure they made that much of an improvement. I'd rather see them redesign the mounting bracket to use after market copper heat sinks. This board was really intriguing, but a pentium-m chip & board is too expensive to run hotter than an athlon 64 venice combo with a decent heat sink. Just my 2 cents.
I wonder if this is the sensor.
Sigh....

As I've noted over and over, the sensor is as accurate as the one on the AOpen 855 board.

I mounted a Zalman 7000 (modded with a 92mm Nexus running at 5V) on the 915 board and the CPU load temps dropped right down to the low 40°C range, exactly the same as is reported by the 855 board using the same CPU and HSF combo.



And as far as this goes:
mc2wheels wrote:61 degrees still seems really hot for the cpu under load.
It's definitely not too hot, but if it worries you, just drop the Vcore down to 1.100V and your load temps will drop to around 50°C, if that makes you feel more comfortable (because the CPU won't care either way).

Posted: Thu Aug 11, 2005 9:10 am
by mc2wheels
I guess I'm just used to my AXP-M 2400+ @1.1v/1.2Ghz with my zalman 7000 with fan running @5v that stays in the high 30's. When I run it overclocked @1.35v/2.0Ghz with the zalman running @7v it runs in the high 40s to 50 range (depending on ambient). Again, I just think it's really a bad design flaw to not allow aftermarket coolers. 61 may not be too hot, but it doesn't seem ideal.

Modified Zalman 7000 Mod

Posted: Tue Aug 16, 2005 11:02 am
by BMW
I bought a 915GMm MB to use in my new HTPC. I didn't like the stock HSF and read that the Zalman 7000 fit, so I got one of those also.

One problem, the 7000 (7000A in my case) interferes with the PCI-E x16 slot. I am using a Sapphire x800 with Arctic Cooling ATI5 Rev 2 cooler. I trimmed about 10 fins that were a problem, and now it looks kind of like a Zalman VF700. I can post a pic if there is interest, but it was not a big deal with a Dremel and a cut-off wheel.

I overclocked the 735 to 2261 MHz (17 x 133), and 1.2V. I fold constantly, so no idle temp, but 49-50C under load. This is in a horribly poor flowing case, so I am very happy. Considering I found the 7000A for $17, it is almost perfect.

So noone else makes the mistake i did, it looks like a VF700 VGA cooler sould fit, but it sits way to close to the motherboard, and interferes with lots of stuff.

Posted: Tue Aug 16, 2005 1:29 pm
by shleepy
Does a 7000 fit without modification (not of the video card, I mean, but of the heatsink)?