Sign in with
Sign up | Sign in
Your question

Asus A7N8X Deluxe, another go

Last response: in CPUs
Share
January 14, 2003 3:50:07 PM

As some of you already know, my A7N8X Deluxe refused FSB's higher than 150MHz when operating in dual-channel sync mode and I had to RMA the board. Now I have got not one, BUT two completely new boards which I am going to try out. Well, if the first board performs well, I'll just return the second without trying it.

It will be interesting to see, how these new boards perform. I really hope they will do as expected, if not, Asus will loose one of it's most loyal ambassadors.

Let the test begin ...

<i>/Copenhagen - Clockspeed will make the difference... in the end</i> :cool:

More about : asus a7n8x deluxe

January 14, 2003 3:55:24 PM

Good Luck!!! I followed your last go & I'll be following this one for sure. I'm a ASUS man myself.

I'm still learning & having fun doing it!!!!!!
January 14, 2003 4:32:09 PM

i am a new asus buyer, but was very dissappointed... but sh!t happens..


Is THG Slightly Slacking??
Related resources
January 14, 2003 4:51:41 PM

Hope this go around works for ya. It's a pain trying to troubleshoot a problem on the working end of things unfortunately.

Keep us posted :) 

Vyr
January 15, 2003 12:20:26 AM

Well I haven't written them off yet. That board will be all right after the Bios catches up to those 333 fsb chips & dual DDR. As they have said on Tom's benches many times, ASUS always gets out a little advantage over the others, almost every time.

I'm still learning & having fun doing it!!!!!!
January 15, 2003 5:39:11 PM

I also eagerly await your results. My 2800+ and A7N8X are at home, waiting for a PSU and some ram for me to try them out...

Marge: Homer, stop picking at your head
Homer: Aw Marge, Im so sweet and tasty!
~~~~~~~~~~~~~~~~~
Ralph Wiggum: I bent my wookie!
January 15, 2003 6:20:35 PM

well you guys think you have it bad... i have had one really bad 2 weeks.

i bought my first a7n8x and it was defective, so i rma'd back to newegg.

got a gigabyte from a local store, that boad didn't work.. the hard drive wasn't being detected.. after a long process of troubleshooting. this board was also defected.. the IDE slots didn't work, therefore it couldn't read the IDE drives..

now i just ordered a new A7N8X off of GoogleGear.com

i hope this works..

oh and one more thing to add to my hell week!

after the first a7n8x setup died.. i go to start up my old one... and that wouldn't start!

turns out the god dam Power supply died.. i mean this is just one hell of a week!

Is THG Slightly Slacking??
a b à CPUs
January 15, 2003 6:31:25 PM

Get Delled!

<font color=blue>You're posting in a forum with class. It may be third class, but it's still class!</font color=blue>
January 15, 2003 6:32:07 PM

That's what happens to electronics while Mercury is in retrograde. :( 


PC Repair-Vol 1:Getting To Know Your PC.
PC Repair-Vol 2:Troubleshooting Your PC.
PC Repair-Vol 3:Having Trouble Troubleshooting Your PC?
PC Repair-Vol 4:Having Trouble Shooting Your PC?
January 15, 2003 7:29:08 PM

Well, I have now almost completed testing the first of the two boards, which is the oldest one of the two boards when judged from the serial number.

<b>Test setup:</b>
Motherboard: ASUS A7N8X Deluxe
BIOS: 1002 Beta 001
CPU: XP2400+, AUICB stepping, wire unlocked
RAM: 2x256MB GeIL PC3500, 6-3-3-2.

<b>Test results:</b>
All tests have been completed with the CPU running as close as possible to the stock frequency for a XP2400+, i.e 2000 MHz.

In <b>single channel</b> mode with memory running in <b>sync</b>, the board was able to run Prime95 stable with <b>210MHz FSB</b>. At 216MHz FSB the board eventually rebooted after running Prime95 for about 10 minutes. I haven't determined the exact stable FSB, but it is somewhere in between 210MHz and 216MHz.

In <b>dual channel</b> mode with memory running in <b>sync</b> things looked a lot worse, almost the same as my original board. The highest stable FSB running Prime95 was <b>158MHz</b>. The original board was stable at 150MHz, so I guess this one is a bit more stable, but certainly not even close to what I was hoping for and it doesn't come close to the results shown in various reviews on the net.

This first board I actually bought myself, because I wouldn't wait for the exchange board to arrive from the ASUS distributor. In the mean time I have received the second board directly from the ASUS distributor. The serial number on the board indicates that it is newer than the one I have just tested. One could maybe hope that some of the stability problems have been ironed out on the new board, but I doubt it. We will see in due time ...


<i>/Copenhagen - Clockspeed will make the difference... in the end</i> :cool:
January 15, 2003 7:42:03 PM

Did the wire unlock trick give you free reign over the multipliers?

Marge: Homer, stop picking at your head
Homer: Aw Marge, Im so sweet and tasty!
~~~~~~~~~~~~~~~~~
Ralph Wiggum: I bent my wookie!
January 15, 2003 8:29:27 PM

Quote:
Did the wire unlock trick give you free reign over the multipliers?

Yes, it unlocks multipliers ranging from 6.0 to 12.5 in 0.5 increments. Using the 1002 Beta 001 BIOS does only give you access to multipliers from 15.0 and upwards (valid for XP2400+).


<i>/Copenhagen - Clockspeed will make the difference... in the end</i> [cool]
January 15, 2003 8:33:03 PM

Maybe you should get rid of that Geil memory, I heard that stuff is crap. I run my A7N8X DLX with dual channel (2x256 Corsair 3500c2) and I'm able to hit a fsb of 207 on 1700tbredB. It runs stable also!

----------------
stop reading this!
January 16, 2003 3:36:14 PM

Quote:
Maybe you should get rid of that Geil memory, I heard that stuff is crap. I run my A7N8X DLX with dual channel (2x256 Corsair 3500c2) and I'm able to hit a fsb of 207 on 1700tbredB. It runs stable also!

No, the RAM is fine. If I run each stick by itself it can run faster than PC3200 RAM at very agressive timings. I can even use a high FSB in dual-channel mode, but only if I run the memory either faster or slower than the FSB frequency. So the dual-channel problem is isolated to synchronous FSB/memory operation.

<i>/Copenhagen - Clockspeed will make the difference... in the end</i> :cool:
January 16, 2003 3:59:36 PM

Did you shove in both sticks when you ran single-channel mode, or just one stick?


PC Repair-Vol 1:Getting To Know Your PC.
PC Repair-Vol 2:Troubleshooting Your PC.
PC Repair-Vol 3:Having Trouble Troubleshooting Your PC?
PC Repair-Vol 4:Having Trouble Shooting Your PC?
January 16, 2003 7:44:09 PM

slvr, I have tried both with one and two stick(s). It doesn't seem to make much of a difference. Dual stick, single channel, sync, works fine up to at least 200MHz FSB with agressive memory timings.


<i>/Copenhagen - Clockspeed will make the difference... in the end</i> :cool:
January 16, 2003 7:44:40 PM

slvr, I have tried both with one and two stick(s). It doesn't seem to make much of a difference. Dual stick, single channel, sync, works fine up to at least 200MHz FSB with aggressive memory timings.


<i>/Copenhagen - Clockspeed will make the difference... in the end</i> :cool:
January 18, 2003 9:42:42 AM

The last board has now been tested. It's a no go. Same problem with all three boards, although not at exactly the same FSB. The first one could run Prime95 stable with a 150 MHz FSB, the second 158 MHz FSB and the third one somewhere in between 150 MHz and 155 MHz, I didn't bother finding the exact value.

All in all, this is rather disappointing. My local dealer has a test-center, and they verified the problems with the first board, so I'm quite sure that my findings are correct. Also both my CPU and RAM can handle high FSB's as long as the memory is not running in sync with CPU bus. I'm not sure that a BIOS update will solve the problem, unless the voltage to the northbridge is software adjustable. Others have been able to increase the FSB by doing a voltage modification and improving cooling of the northbridge. I could also try this, but the problem is that the warranty is void, so I'm a bit reluctant. I feel that Asus should have delivered a board capable of running the new generation Thoroughbred B's with 333MHz bus, in dual-channel mode sync, right out of the box without any need for modification.

Too bad, I really liked the huge amount of features on this board, but it seems to come at a performance penalty. If I can't get a guarantee that the problem will be solved by Asus, I'll demand a refund and then go buy the Epox nForce2 board instead.

<i>/Copenhagen - Clockspeed will make the difference... in the end</i> :cool:
January 19, 2003 9:23:37 PM

how the hell are you getting fsb of over 200? i have an asus a7n8x and the max in bios is 200, what bios revision are you using? thats insane.
also using dual channel i can do about 180 fsb (xms 3200 running sync and aggressive).
thats w/auto on the agp (meaning it gets ocd with the fsb) and i dont know if on the asus u can change the pci multi i dont think so though, if so that could help.

The first LAN party I went to was at a PETA convention. They booted me when I shot a crab in HL!
January 19, 2003 9:54:26 PM

I'm using the latest Beta BIOS from Asus, 1002 Beta 3.

It allows for FSB's all the way up to 250 MHz as opposed to the standard 211 MHz, but that's pure academic with the HUGE stability problems this board has in dual-channel/ sync memory mode.

I can also do 180 MHZ FSB, but it's not stable at all. Try Prime95 and 3DMark2001SE and run them for an extended period of time. Then tell whether you are still stable with 180MHz FSB.

What CPU (model and stepping) do you have ?


<i>/Copenhagen - Clockspeed will make the difference... in the end</i> :cool:
January 19, 2003 9:59:36 PM

xp 2700 standard multiplier.
i have 3dmark2001 and it ran fine but i only did it one time.

The first LAN party I went to was at a PETA convention. They booted me when I shot a crab in HL!
January 19, 2003 11:15:18 PM

Which memory slots are you using ?

There's one black, followed closely by a blue slot and then further away there's a second blue slot.


<i>/Copenhagen - Clockspeed will make the difference... in the end</i> :cool:
January 20, 2003 1:34:43 PM

Hahaha, no Sh*t, how about taking into account that this is a toms hardware board and the fact that ASUS is almost always in the lead in all mobo reviews and since people here actually base their purchases on these reviews, the ratio of asus owners to msi owners is quite disproportional (asus is highly regarded on all hardware review sites). Also, judging a mobos stability only by its manufacturer is rather stupid. You can take the most stable mobo and give it crappy ram, or incompatible peripherals, or a dumbass user - and you get a piece of poop stability-wise (dont even get me started on OS maintenance).

Damn it, i want to use htm here.
Man....damn....may rambus take your soul.
January 20, 2003 1:53:49 PM

First of all. Im assuming you used the same power supply that died for the first board. Did you stop to consider your power source at all? I mean, a power supply that is dying can damage your board at worst and make it unstable atleast. Also, a good high quality powersupply is crucial for stable performance at high speeds. You want a clean, regulated voltage. Did you consider the actual electrical wiring of your home? Sometimes people run too many appliances on the same fuse, and then wonder why their computer is not doing so well. That can actually damage your power supply by the way, especially a crappy one. MAybe you need a line conditioner...

Geil is NOT OK, NOT FINE. You should test the mobo with a different brand. You are here asking for help, so listen to what you are being told. GEIL has a reputation for crappiness. Im not going to waste my time going into details, it has been very well documented. Just read some forums about it. Until you try different ram, dont be RMAing your boards left and right. The fact that you tried several boards with the same ram and none can exceed a certain speed should make it quite obvious that its not the MOBOs fault. Especially since people allready run much more aggressive timings with better ram.

How did you come to the conclusion that your ram is FINE anyway? Did you test it?

Damn it, i want to use htm here.
Man....damn....may rambus take your soul.
January 20, 2003 4:26:19 PM

I would agree, try using different RAM in there and see what happens. Corsair has been successful for me and a few others it seems. I'm running dual channel Corsair 256 x2 PC3200 in synch at 191 FSB. 3 different boards in a row with the same limitiations seems highly unlikely.

Ohh, and as far as timings go, I'm running 2/2/2/5.

Vyr
January 20, 2003 6:01:29 PM

stoochie, you may want to press reply to one of my posts, instead of someone elses (Rob), after all, judging from your reply it looks like your mail is mainly directed to me.

The specific memory I have, GeIL PC3500 Ultra, is, I repeat, perfectly fine. It's specified to be able to run 433 MHz with 6-3-3-2 timings, which I have verified, using a <b>single</b> stick in <b>non-sync</b> mode.

My first board was able to reach a FSB of only 150MHz in dual-channel mode with synchronized memory frequency (2x150 MHz) without losing stability. Using a x12.5 multiplier my stock 2000MHz XP2400+ was then running only 1875 MHz. My local dealer verified the FSB limitation using an XP2600+ with DDR333 bus and some other type RAM (I don't know the brand and specifications).

There is absolutely no doubt in my mind, that the ASUS A7N8X Deluxe, even with it's latest BIOS, has serious FSB problems in dual-channel sync mode. Search the WEB, and you will find an increasingly number of people experiencing the exact same problems as Rob and I have had. It certainly looks like ASUS rushed this product to market, just in time for x-mas.

Regarding power-supply: I even bought an Antec 480W TruePower, in an attempt to attain stability, but it didn't make a difference AT ALL. I'll tell you, this board has been a very expensive and time-consuming encounter, adding all the hours debugging and switching of various components. I have e-mail'ed ASUS long time ago, but they never bothered answering me. Lots of people are very disappointed and like me, they have also used huge amounts of time on this unfinished product. A few people seems to be able to reach a reasonably high FSB, around 190MHz, by adding active cooling to the Northbridge. I think that maybe a higher voltage to the Northbridge combined with improved cooling would allow for a higher FSB. There is one modded BIOS around which allows increasing the chip voltage with 0.2V, but the underlying software/hardware doesn't implement it yet. Probably ASUS is a bit reluctant to allow this because the factory Northbridge cooling is insufficient.

Finally: People are wondering why ASUS didn't allow for unlocking of Thoughbred B's in their BIOS like Epox and others; I'll tell you why: It would have made it evident (like in crystal clear) for even more people, that the board simply can't cope with the high FSB's at all and this would have caused just even more complaints. Actually the last BIOS from ASUS in fact unlocks the multipliers, BUT ONLY THE HIGHER ONES, LOL !!!!


<i>/Copenhagen - Clockspeed will make the difference... in the end</i> :cool:
January 20, 2003 8:51:04 PM

Quote:
would agree, try using different RAM in there and see what happens. Corsair has been successful for me and a few others it seems. I'm running dual channel Corsair 256 x2 PC3200 in synch at 191 FSB. 3 different boards in a row with the same limitiations seems highly unlikely.

Yes, it do seems like Corsair stands out from the rest of the pack, I noticed that too. But that stuff is really expensive and can hardly be justified. It lowers the performance/price ratio quite a bit.

I'm puzzled. One the one hand my 256MB GeIL PC3500 Ultra sticks performs fine one stick at a time. But put two together and especially in dual channel sync mode, that's a completely different story. Whether I should blame GeIL or ASUS is hard for me to decide. If just I was able to try out a pair of Corsair CAS2.

<i>/Copenhagen - Clockspeed will make the difference... in the end</i> :cool:
January 21, 2003 12:56:55 AM

im using the two blue ones

The first LAN party I went to was at a PETA convention. They booted me when I shot a crab in HL!
January 21, 2003 7:22:31 AM

The performance of GEIL/OCZ can be easily explained. The "Hand Selected" samples that they test for high speed is probably not tested for dual channel synchrous operation. So although they are quite exceptional at providing the speed, I bet the difference in sample origins and the fact that they arent really the cream of the crop when it comes down to selection - prevents their operation at agressive timings and still stay in sync. Corsair is a high quality memory from the ground up and thats what you pay for. And unlike GEIL they arent resold so you end up paying a reasonable price (look on the net, i got my 512 DDR 400 pc3500 corsair stick for 200bucks.) Another manufacturer to look into is samsung, they have pretty good quality, im not sure how they measure up to corsair but they are definitelly higher quality then GEIL/OCZ

Damn it, i want to use htm here.
Man....damn....may rambus take your soul. <P ID="edit"><FONT SIZE=-1><EM>Edited by stoochie on 01/21/03 04:27 AM.</EM></FONT></P>
January 21, 2003 8:07:04 PM

I think GeIL PC3500 Ultra performs great when using a single stick. I bought 2 sticks, but they were not handpicked for dual-channel operation.

In seems like nForce2 boards are VERY sensitive regarding the used RAM. x-bit-labs just tested a Chaintech 7NJS using 2 x 256MB DDR400 CAS2 Corsair XMS3200 memory. They couldn't take the FSB past 166MHz without problems. Maybe this is why Corsair has started to introduce the TWINX512-3200LL which is a matched pair of memory modules which has been verified to work on an ASUS A7N8X in dual-channel mode. This indicates to me that the nForce2 boards are very picky about the memory used, the Epox implementaion being a possible exception. This is a MAJOR weakness in the nForce2 design or implementation, because it makes the complete solution very expensive and almost impractical.

After reflashing the official BIOS 1001.C and playing around with the modules in the different banks (1-3, 2-3, 3-2, 3-1) and lowering the CAS value to 2.5, I'm now stable at 186MHz FSB, dual-channel mode and sync memory frequency. Memory settings are now: 5-2-2-2.5 which is of course not so impressive. Placing the "weakest" module in slot 1 and the best module in slot 3 made a big difference as did running CAS 2.5 instead of CAS 2. No doubt that a single 512 MB module running a 200MHz FSB with 5-2-2-2 timings would perform better; so much for the highly regarded dual-channel operation of the nForce2. Only if the board had integrated graphics it would benefit from the increased memory bandwidth, but it doesn't ...


<font color=green><b>Update:</b> I tried to increase the FSB just a bit more: Now I'm at 188MHz FSB with a x12.5 multiplier, i.e my XP2400+ (2000MHz) is now running 2350MHz. I've not tested it in depth, but it's completed a full Prime95 run (992 Test). During the test Asus Probe Ver. 2.19.05 reported:

CPU: 32 C/ 89 F
MB: 30 C/ 86 F

This is with the case open and a low noise Pabst 92mm fan blowing on the Northbridge and RAM. </font color=green>

<i>/Copenhagen - Clockspeed will make the difference... in the end</i> :cool:
<P ID="edit"><FONT SIZE=-1><EM>Edited by Copenhagen on 01/21/03 06:10 PM.</EM></FONT></P>
January 22, 2003 10:38:19 AM

So are you still expecting to get the Epox board any soon?

--
This post is brought to you by Eden, on a Via Eden, in the garden of Eden. :smile:
January 22, 2003 3:28:02 PM

Quote:
So are you still expecting to get the Epox board any soon?

It would certainly be interesting to find out if the Epox EP-8RDA+ is less picky about the used memory modules.

I've not decided yet whether I should keep the memory and sell the board or keep the board and get some better memory.


<i>/Copenhagen - Clockspeed will make the difference... in the end</i> :cool:
January 22, 2003 3:34:00 PM

The question is...well changing ur memory allow u to use the dual mem arch. in th asus board...if it does than stick with the asus cuz than u can have the best board with the most features........

Therefore i say ditch the mem. and get new mem.....

<A HREF="http://www.anandtech.com/mysystemrig.html?id=13597" target="_new">-MeTaL RoCkEr</A>
January 22, 2003 3:43:10 PM

Tough decision after all you've done with the ASUS.

I'm still learning & having fun doing it!!!!!!
January 22, 2003 3:55:49 PM

Yes, A7N8X Deluxe has more features and is well designed, but it lacks in the overclocking department. Epox has tuned everything towards peak performance, such as a 3-phase voltage regulator instead of 2-phase, and the board layout is not so nice, but again designed for max. performance. It really depends whether you need S-ATA RAID and dual NIC.

<b>ASUS A7N8X Deluxe</b>
Max. CPU voltage: 1.85V
Max. AGP voltage: 1.7V
Max. DDR voltage: 2.8V

<b>Epox EP-8RDA+</b>
Max. CPU voltage: 2.20V
Max. AGP voltage: 1.8V
Max. DDR voltage: 2.9V

At the moment I'm not using dual NIC nor S-ATA RAID, so if I could redo my purchase, I would go for the Epox. My watercooling gear is not challenged at all, running Prime95 at 2350MHz and 1.85V causes almost no temperature increase, only 32 C/ 89 F, cool to the touch.

<i>/Copenhagen - Clockspeed will make the difference... in the end</i> :cool:
January 22, 2003 4:09:23 PM

Awesome....well than i would say than now is the time for u to make ur decission......if OC'ing is all ur caring about than get that Epox board get rid of the 2 ASUS Boards u have now.......and ur set.....that..or get rid of 1 of the asus boards and keep the other for future use or even later on bios development...who knows.....

<A HREF="http://www.anandtech.com/mysystemrig.html?id=13597" target="_new">-MeTaL RoCkEr</A>
!