Motherboards that support a 2500+ mobile, and Zalman 7000?
Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee
-
- Posts: 11
- Joined: Tue Apr 20, 2004 2:01 pm
- Location: Atlanta, GA
- Contact:
Motherboards that support a 2500+ mobile, and Zalman 7000?
Greetings!
I recently picked up a Barton 2500+ mobile last night for about $89. Not the bet price, but it was local, so I dealt with it.
This was going to replace my XP1600+ I was using in an Asus A7V266 motherboard. I had read online that this board supported the 2500+ mobile in its latest firmware revision -- but I was wrong, it was the A7V266-E that supported it, not the plain 266.
So, now I'm on a mission:
Since I've got my shiny new Barty mobile in my motherboard, and it's only able to run at 800MHz (any attempt at raising the multiplier to the correct level causes the system to hang. Motherboard compatibility issue I'm guessing), I need a new motherboard.
But, I have a few critera:
1) It must be able to support the Barton 2500+ mobile CPU.
2) It must be able to mount a Zalman 7000Cu.
3) (and this is the kicker) it has to fit into a Lian-Li 6070, which doesnt have a whole lot of room between the power supply and the top of the motherboard. Im at work now, so I cant check inside my case to get a better idea of how much room is up there -- butit's not a whole lot if I recall correctly. Can anybody prove me wrong?
Thanks!
I noticed the Abit KT7A seems to support all my criteria. Anybody have experience with this board? The last time I used an Abit board was a dual celery BP6, hehe. I prefer Asus, but any maor brand I wouldnt be opposed to.
Thanks for the help!
I recently picked up a Barton 2500+ mobile last night for about $89. Not the bet price, but it was local, so I dealt with it.
This was going to replace my XP1600+ I was using in an Asus A7V266 motherboard. I had read online that this board supported the 2500+ mobile in its latest firmware revision -- but I was wrong, it was the A7V266-E that supported it, not the plain 266.
So, now I'm on a mission:
Since I've got my shiny new Barty mobile in my motherboard, and it's only able to run at 800MHz (any attempt at raising the multiplier to the correct level causes the system to hang. Motherboard compatibility issue I'm guessing), I need a new motherboard.
But, I have a few critera:
1) It must be able to support the Barton 2500+ mobile CPU.
2) It must be able to mount a Zalman 7000Cu.
3) (and this is the kicker) it has to fit into a Lian-Li 6070, which doesnt have a whole lot of room between the power supply and the top of the motherboard. Im at work now, so I cant check inside my case to get a better idea of how much room is up there -- butit's not a whole lot if I recall correctly. Can anybody prove me wrong?
Thanks!
I noticed the Abit KT7A seems to support all my criteria. Anybody have experience with this board? The last time I used an Abit board was a dual celery BP6, hehe. I prefer Asus, but any maor brand I wouldnt be opposed to.
Thanks for the help!
-
- SPCR Reviewer
- Posts: 2696
- Joined: Thu Dec 11, 2003 9:53 pm
- Location: Scarsdale, NY
- Contact:
KT7A would rate as, "Ancient," and probably still does not support Mobile Barton cores; my recommendation for you would be the AN7, and then swapping out the stock northbridge cooler for something passive. This is exactly what I have done in my machine, Gamma Two; check out that post in the gallery for images of the build. My case places the PSU literally right up to the board, and I didn't have a problem fitting the Thermalright SP-97 on there with a Panaflow M1B, which comes out to 99mm in width overall. Your Zalman CNPS7000A-Cu/AlCu comes out to 109mm in diameter, making it 5mm wider in each direction; judging by the looks of it, it will still fit a CNPS7000 just fine, and the AN7 accepts any heatsink on the market that is designed for Socket A, either through-board mounting, or socket-mounting.
As you can see in the gallery, I, too, am running a Mobile Barton core, only of the 2400+ grade, and it works fine on my AN7. The only thing that may appear strange is that Windows will indicate, "Unkown CPU Type," but it works fine, and if you want to ensure full SSE, MMX and 3D-Now! support, just manually install the AMD K7 driver for the CPU within Windows' Device Manager as I did.
-Ed
As you can see in the gallery, I, too, am running a Mobile Barton core, only of the 2400+ grade, and it works fine on my AN7. The only thing that may appear strange is that Windows will indicate, "Unkown CPU Type," but it works fine, and if you want to ensure full SSE, MMX and 3D-Now! support, just manually install the AMD K7 driver for the CPU within Windows' Device Manager as I did.
-Ed
Edward Ng - hmmm, the CPU detection is kinda weird
Mine is normal old unlocked Barton:
...and Epox 8RDA+ detecting it pretty well... Im about to get the mobile (hope it comming...) pretty soon, so I have to be prepared for "unknown CPU" as well?
I tought the Abit AN7 should be better, not even bios update fix this glitch?
Mine is normal old unlocked Barton:
...and Epox 8RDA+ detecting it pretty well... Im about to get the mobile (hope it comming...) pretty soon, so I have to be prepared for "unknown CPU" as well?
I tought the Abit AN7 should be better, not even bios update fix this glitch?
-
- Posts: 16
- Joined: Mon Mar 03, 2003 2:28 pm
- Location: London
-
- SPCR Reviewer
- Posts: 2696
- Joined: Thu Dec 11, 2003 9:53 pm
- Location: Scarsdale, NY
- Contact:
I have not yet come across a desktop board, personally, that will detect, specifically, the Mobile Barton...
Actually, the AN7 and the FN41V3 (Shuttle SN41G2V2) both detect it as a Mobile Athlon XP when left to their own vices (only 600MHz for my CPU!), but when set for higher clockrates, they switch to Unknown CPU Type.
This is the consequence of using a mobile chip on a desktop platform; simply have to get used to it. As I said, it works fine.
-Ed
Actually, the AN7 and the FN41V3 (Shuttle SN41G2V2) both detect it as a Mobile Athlon XP when left to their own vices (only 600MHz for my CPU!), but when set for higher clockrates, they switch to Unknown CPU Type.
This is the consequence of using a mobile chip on a desktop platform; simply have to get used to it. As I said, it works fine.
-Ed
-
- SPCR Reviewer
- Posts: 2696
- Joined: Thu Dec 11, 2003 9:53 pm
- Location: Scarsdale, NY
- Contact:
ONLY the revision 1.2 and 2.0 NF7 and NF7-S have the mounting holes, so be careful! Revision 1.0 boards did not! Of course, it would be difficult to still find a board of that revision, but if you're shopping auction sites or looking at refurb or old stock, you need to watch out for that.
All AN7 have the mounting holes.
-Ed
All AN7 have the mounting holes.
-Ed
However, some of the NF-7 series boards (NF-7S I think) as well as AN7 have issues with CNPS-7000 - there's a capacitor in the way. You have to either file down the heatsink to clear the capacitor, or take a soldering iron to the board, re-mounting the capacitor horizontally with some longer wires, and destroying the warranty.
Last edited by Gholam on Fri May 14, 2004 1:54 pm, edited 1 time in total.
Unfortunately the NF7 has the CPU socket right up against the edge of the board near the PSU. If you have a case where the PSU is almost touching the motherboard (such as my Ever Case 4252, for example), there's no way you're going to fit a Zalman CNPS-7000 on there, as the original poster desires.Rusty075 wrote:You can add the Abit NF7 series to the list of mobo possibilities. XP-M works just fine, and it has the HSF mounting holes.
Other than that the NF7-S is one of the best, most stable Nforce2 boards on the market, once you've replaced the NB fan with a Zalman NB-47J.
-
- SPCR Reviewer
- Posts: 2696
- Joined: Thu Dec 11, 2003 9:53 pm
- Location: Scarsdale, NY
- Contact:
Metaluna if you look at the link I gave him, it shows images of my system, and you will see that in my system, the PSU runs extremely close to the board. The heatsink I'm using is 10mm less wide in general, which means 5mm narrower on each side; looking at the clearance in my system, there's a good 8mm clearance between my heatsink and the PSU, so the CNPS7000 would still fit.
The mainboard is about 6mm or so away from the PSU in my case; if the mainboard is closer than that in your case, well, I really don't know what to say; whoever designed the case must be crazy to make the tolerances that close!
It's simple; look at the images of my system in that link from the gallery above, and if your spacing looks like mine, the CNPS7000 will still fit, just extremely close. Remember that my sink has metal clips holding the fan on, and they extend even farther from the sink, yet it still fits in my system (AN7).
-Ed
The mainboard is about 6mm or so away from the PSU in my case; if the mainboard is closer than that in your case, well, I really don't know what to say; whoever designed the case must be crazy to make the tolerances that close!
It's simple; look at the images of my system in that link from the gallery above, and if your spacing looks like mine, the CNPS7000 will still fit, just extremely close. Remember that my sink has metal clips holding the fan on, and they extend even farther from the sink, yet it still fits in my system (AN7).
-Ed
Well, Edward Ng, glad to hear yours fit
There is a list of compatible mobos on the Zalman site:
http://www.zalman.co.kr/product/cooler/ ... t_eng.html
...and there are even mentioned things like "Power Supply must not interfere with the heatsink", or "Clip must be cut off to avoid interference" so they even listed mobos, that fit with minot tweaks
However in case of Abit AN7 mobo is marked as Incompatible on their site I did not having the mobo (but I do have CNPS7000Cu), so I can't judge what the hell this means, however even the list is kinda precise, it lack mobo revisions, so perhaps there are AN7 revision, that make it incompatible with CNPS7000Cu?
OR you did some tweaking?
When looking at some your pics it seems that it should fit on the AN7, even the NB cooler might get into a little trouble (as you already experienced for yourself), hmmm
Weird.
For the mobile Barton recognized as unknown CPU, I currently waiting (like for a miracle) for mine, so we see what Epox 8RDA+ say about it
There is a list of compatible mobos on the Zalman site:
http://www.zalman.co.kr/product/cooler/ ... t_eng.html
...and there are even mentioned things like "Power Supply must not interfere with the heatsink", or "Clip must be cut off to avoid interference" so they even listed mobos, that fit with minot tweaks
However in case of Abit AN7 mobo is marked as Incompatible on their site I did not having the mobo (but I do have CNPS7000Cu), so I can't judge what the hell this means, however even the list is kinda precise, it lack mobo revisions, so perhaps there are AN7 revision, that make it incompatible with CNPS7000Cu?
OR you did some tweaking?
When looking at some your pics it seems that it should fit on the AN7, even the NB cooler might get into a little trouble (as you already experienced for yourself), hmmm
Weird.
For the mobile Barton recognized as unknown CPU, I currently waiting (like for a miracle) for mine, so we see what Epox 8RDA+ say about it
I would guess the AN7 is marked "incompatible" for the same reason as the NF7 series: you need to reverse the plain aluminum and blue anodized mounting brackets from the positions given in Zalman's CNPS7000 mounting instructions. This is necessary to clear the closest capacitor to the socket, and shifts the HSF approx 6mm towards the locking lever side of the CPU socket. There is still plenty of contact between the CPU core and the HSF. You also must cut off one end of the Zalman mounting clip (as shown in the instructions). I successfully mounted a CNPS 7000 to my NF7-S in a Sonata (another case with a PSU close to the mobo edge) this way.trodas wrote:However in case of Abit AN7 mobo is marked as Incompatible on their site I did not having the mobo (but I do have CNPS7000Cu), so I can't judge what the hell this means, however even the list is kinda precise, it lack mobo revisions, so perhaps there are AN7 revision, that make it incompatible with CNPS7000Cu?
OR you did some tweaking?
-
- SPCR Reviewer
- Posts: 2696
- Joined: Thu Dec 11, 2003 9:53 pm
- Location: Scarsdale, NY
- Contact:
Ah, I did not see their list of compatibility; remember that I am not using the CNPS7000, I am using the SP-97. I only guessed it would fit by comparing the physical measurements for both heatsinks.
In that case, I can only recommend to use an SP-97 with the AN7, as I have done, but as you know, that heatsink is more expensive than CNPS7000.
-Ed
In that case, I can only recommend to use an SP-97 with the AN7, as I have done, but as you know, that heatsink is more expensive than CNPS7000.
-Ed
-
- SPCR Reviewer
- Posts: 2696
- Joined: Thu Dec 11, 2003 9:53 pm
- Location: Scarsdale, NY
- Contact:
I have not personally tried the CNPS7000 for myself, so I cannot vouch for whether or not it can outperform the SP-97 (although MikeC's comparison tends to prove the SP-97 superior) and I am pretty confident that the SP-97 will run quieter than the CNPS7000 when aiming for the same thermal performance (i.e. CNPS7000 fan at 12volts vs. SP-97 with a quieter fan); I'm pretty sure you've seen all three of my machines in the Gallery (Sigma One, which is a sticky, Alpha Three and Gamma Two), and I honestly am not very sure that I can get that level of performance (thermal combined with acoustic) out of Zalman's big bad boy.
However, as I said, I have not yet directly compared them like MikeC has. The fact that SP-94/7 are through-board mount, don't stick up as far from the board (applies less torque), weighs much less than the pure copper version (the model that I believe has a chance at performing as well as SP-94, 773g vs. 540 g., and the AlCu, which is not as good, does weigh less than both, around 453) and fits more boards.
Simply put, the CNPS7000A-AlCu is a great bargain, providing the lowest weight and cost with about 85% the thermal performance of the two copper-only sinks, then the SP-94 is a better choice over the CNPS7000A-Cu because it's lighter, fits more boards, performance at least as well and puts less stress on your mainboard.
-Ed
However, as I said, I have not yet directly compared them like MikeC has. The fact that SP-94/7 are through-board mount, don't stick up as far from the board (applies less torque), weighs much less than the pure copper version (the model that I believe has a chance at performing as well as SP-94, 773g vs. 540 g., and the AlCu, which is not as good, does weigh less than both, around 453) and fits more boards.
Simply put, the CNPS7000A-AlCu is a great bargain, providing the lowest weight and cost with about 85% the thermal performance of the two copper-only sinks, then the SP-94 is a better choice over the CNPS7000A-Cu because it's lighter, fits more boards, performance at least as well and puts less stress on your mainboard.
-Ed
-
- Posts: 11
- Joined: Tue Apr 20, 2004 2:01 pm
- Location: Atlanta, GA
- Contact:
I should probably follow up my post!
I eventually settled with an Asus A7N8X-X. It doesn't have everything I want, but the price was right, and reviews seem good. So far the jury is still out on whether or not I like it.
Yes, it does take the 2500+ mobile. It took me a while to figure out the correct multiplier and external speed (166 x 11 is correct, btw, for stock 1.833GHz). Although, after changing these settings, the CPU type is not recognized by the motherboard on bootup. No big deal.
Also, this board will not volt lower than 1.575V. This is a little higher than what the 2500+ M requires for stock speed, 1.45V. But I dont think this is much of an issue, given that the 2500+ mobile chip is labelled mobile because it's able to maintain its speed with a lower Vcore.
To make myself feel better about the higher voltage, I went ahead and put the multiplier at 12X for a nice and even 2000MHz. I bought this chip mainly because of its low 45W heat output, and my case is fairly restrictive (Lian Li 6070B). Though I have cut out the rear fan grill and dremeled out the restrictive holes on the breather plate underneath. Next step is new fans, and perhaps adding a second 80mm exhaust, as there appears to be room under the stock 80mm back there.
My Zalman 7000Cu cooler clears the bottom of the power supply in this case by about an inch. I was curious to know why Zalman had the A7V8X-X listed as INCOMPATIBLE with their cooler and the A7N8X-X as compatable only if your power supply didn't interfere -- looking at the A7V8X-X, it seemed like it had a whole lot more room around the 462 socket than the A7N8X-X did, and the mounting holes were present. I was pretty sure the cooler would fit, but I didn't buy it because I trusted Zalman's research, there -- I didn't want to get home later and figure out the reason Zalman had it listed as incompatable.
The Zalman 7000Cu clears the capacitors just south of the 462 socket on the A7N8X-X by about 2mm. I was worried about this at first, but it's no problem.
Here's a question for all you techies, and I might post this as a new thread:
In my recent adventures in motherboard and CPU purchasing, I'm confused about what exactly the FSB is. My 2500+ Barton is labelled as having a stock FSB of 266MHz, and my RAM is PC2100 DDR, which should match. However, my BIOS reports my ram operating @ 133MHz, and the only options for the External Speed of the chip are 100, 133, 166, and 200MHz. Am I missing something? The nForce2 400 chipset on this board should support up to a 400MHz FSB, correct?
-db.
I eventually settled with an Asus A7N8X-X. It doesn't have everything I want, but the price was right, and reviews seem good. So far the jury is still out on whether or not I like it.
Yes, it does take the 2500+ mobile. It took me a while to figure out the correct multiplier and external speed (166 x 11 is correct, btw, for stock 1.833GHz). Although, after changing these settings, the CPU type is not recognized by the motherboard on bootup. No big deal.
Also, this board will not volt lower than 1.575V. This is a little higher than what the 2500+ M requires for stock speed, 1.45V. But I dont think this is much of an issue, given that the 2500+ mobile chip is labelled mobile because it's able to maintain its speed with a lower Vcore.
To make myself feel better about the higher voltage, I went ahead and put the multiplier at 12X for a nice and even 2000MHz. I bought this chip mainly because of its low 45W heat output, and my case is fairly restrictive (Lian Li 6070B). Though I have cut out the rear fan grill and dremeled out the restrictive holes on the breather plate underneath. Next step is new fans, and perhaps adding a second 80mm exhaust, as there appears to be room under the stock 80mm back there.
My Zalman 7000Cu cooler clears the bottom of the power supply in this case by about an inch. I was curious to know why Zalman had the A7V8X-X listed as INCOMPATIBLE with their cooler and the A7N8X-X as compatable only if your power supply didn't interfere -- looking at the A7V8X-X, it seemed like it had a whole lot more room around the 462 socket than the A7N8X-X did, and the mounting holes were present. I was pretty sure the cooler would fit, but I didn't buy it because I trusted Zalman's research, there -- I didn't want to get home later and figure out the reason Zalman had it listed as incompatable.
The Zalman 7000Cu clears the capacitors just south of the 462 socket on the A7N8X-X by about 2mm. I was worried about this at first, but it's no problem.
Here's a question for all you techies, and I might post this as a new thread:
In my recent adventures in motherboard and CPU purchasing, I'm confused about what exactly the FSB is. My 2500+ Barton is labelled as having a stock FSB of 266MHz, and my RAM is PC2100 DDR, which should match. However, my BIOS reports my ram operating @ 133MHz, and the only options for the External Speed of the chip are 100, 133, 166, and 200MHz. Am I missing something? The nForce2 400 chipset on this board should support up to a 400MHz FSB, correct?
-db.
The A7N8X does support 400MHz fsb. please someone correct me if I'm wrong, but I think DDR ram opperates at an actual clock speed of half of its rating. So 333MHz ddr actually operates at 166MHz. But ddr is able to transport twice the data (compared to old SDRAM) cause it can use both the rising and falling edge of the clock signal - I think SDR only uses one edge. I hope that made some sence
Hint: DDR = double data rate
200 MHz x 2 = 400MHz
Hint: DDR = double data rate
200 MHz x 2 = 400MHz
the_smell - yes, you are pretty right - DDR values are twice the normal ones - tought the clock remain the same. Sort of marketing bulls*it, if you ask me, since replacing SDRAM to DDRs bring only 5-6fps more in Q3 engine timedemo test
No big deal.
Mine memory/FSB clock is 200Mhz and is reported as 400Mhz "Bus speed" (x2):
Things get more confusing soon, tought.
DDR2 and DDR3 memories comming. Mine FX5700U using DDR2, and their real clock are 4x smaller the the boosted one DDR3 takes this "great clock thing" even futher, boosting 8x the memory actual clock as rating
Comparing the DDR2 and DDR3 special edition of the same FX5700U i own show, that the differences between DDR2 and DDR3 are next to none (okay, DDR3 "wins" by about 1-2fps ) and companies probably will use then for two main reasons, over DDR2 - less voltage/heat and marketing boosting
After all, multiplying the clock 8x show very nice Mhz
Mine memory is capable of stable operation at 970Mhz, and can go for a while to 1024Mhz, witch is 1Ghz - pretty nice, right?
The sad truth is, that the memory clock is operating at 256Mhz then, and therefore this is simply no big deal
While the much faster block data transfer sure rules, the inability of RAS/CAS selection performing faster probably limit the speeds of GFX cards notably
/end of rant
No big deal.
Mine memory/FSB clock is 200Mhz and is reported as 400Mhz "Bus speed" (x2):
Things get more confusing soon, tought.
DDR2 and DDR3 memories comming. Mine FX5700U using DDR2, and their real clock are 4x smaller the the boosted one DDR3 takes this "great clock thing" even futher, boosting 8x the memory actual clock as rating
Comparing the DDR2 and DDR3 special edition of the same FX5700U i own show, that the differences between DDR2 and DDR3 are next to none (okay, DDR3 "wins" by about 1-2fps ) and companies probably will use then for two main reasons, over DDR2 - less voltage/heat and marketing boosting
After all, multiplying the clock 8x show very nice Mhz
Mine memory is capable of stable operation at 970Mhz, and can go for a while to 1024Mhz, witch is 1Ghz - pretty nice, right?
The sad truth is, that the memory clock is operating at 256Mhz then, and therefore this is simply no big deal
While the much faster block data transfer sure rules, the inability of RAS/CAS selection performing faster probably limit the speeds of GFX cards notably
/end of rant