*** Official AOpen AX4C/Max (875P) Thread ***

[shoman94]

room temp=22c
system temp=27c
cpu idle=29.5c
prime95 load=46c

Actual voltage is less than bios....but either way 1.700 is my limit for the cpu....
Black slots allowed me to go higher. More stable....

 

[duncan888]

I've noticed that Aopen have taken the 1.04a beta bios off the download page. Have they finally realised that its got a major problem with it!! There taken there time to release the next bios, every other motherboard seem to have updates apart from this one!

 

[shoman94]

My guess would be there about to release a new one. You can still get the r1.04a bios thru the ftp.aopen.com site though.

 

[Archangel07]

Sorry about the delay in the response, but I have been busy. If you looking for the oc tests done on the AX$C MAx you can find them here: http://english.aopen.com.tw/tech/report/overclk/mb/AX4C-Max_0.htm

It appears that the tests were done with the 1.04 bios version. Interesting results with standard cooling.

 

[Archangel07]

As an aside, its my opinion that the 1.04 may be the end to the bios revisions as Aopen was able to reach the 250 fsb sweet spot with air cooling. But I could be wrong, and I hope I am as I would like to see the voltage settings to go a little higher

 

[StuPidaso]

OK..have built computers for a few years now but just made the leap from a 900 Celeron to my Current system you can see in my signature. My question is what is the difference between cpu temp and cpu kernel temp? Also if I run SANDRA it displays the cpu kernal as "aux/power" temp and says it is too high. Thanks guys.

 

[Archangel07]

I am not sure what the kernal temp is, but from a review of the manual and monitoring during my OC adventures it does not appear to rise above 69c. Sandra is good for some things, but stress testing your puter is not one of them. Download prime95 and monitor your temps through Silentek included with software bundle or MBM, btw don't try to run MBM and ST at same time, you won't like the noise

Well after playing with the AX4C Max over this last week I have reached a 252 FSB 2.8c@3.53 (rounded up, as there appears to some discrepancy between bios and WCPU readings). Very stable at this speed, as prime95 ran for approximately 45 minutes without a glitch. Was disappointed in early trials with 1:1 ratio as I could not get the board to OC past 215, and thus was not able to reach the 433 speed of my mem. My PCMark scores are in the 9ks and I am very pleased with the board.

I did not have problems with cold boots using the 1.4a beta, in fact the 1.4 gave me cmos errors on reboot and kept resetting my vcore to default.

Anyway, I fully expect to reach in between 255-260 without trouble, but I like my stability so I am slowing down the process to let the puter adust 1 mhz at a time.

Here is my config:

AX4C Max BIOS 1.04a
2.8c@3.53 252 FSB
2x512 Geil Dragon 3500
3XSCSI U160 on Mylex 170 Raid controller
Be Cooling Slit edge WB CPU; Gemini Chip set WB; Gemini G4 WB; Via 1300 1/2 system
Ti4400 (OC'ed to hell and back)

Shoman, what sys temps are you getting during prime run. I had to mod the WB in order to get it to mount and I need a reference point for comparison. I am topping out at 40c on cpu temp at load, but my sys temp is going as high a 37c, to little of a difference.

 

[jbur293]

Hi all!

I just picked up an Aopen AX4C Max motherboard and have it up and running. Here's my setup

Aopen AX4C Max - Bios version 1.04a
Intel Pentium 4C 2.4GHz (overclocked to 3.0GHz)
Memory - Geil Golden Dragon 3200DC (256MBx2) 2-3-3-6 running at 1.33 ratio (Stable)
Antec 380w True Power
Western Digital 80GB SE
Leadtek Geforce3 Ti200

I have overclocked the processor to 3.12GHz with 1.33 ratio and would boot into WinXP fine but would not run the Prime95 torture test. I could also run 3.0GHz at 1.66 ratio but could not run the Prime95 torture test. I'm using the back two slots for dual channel as suggested above. Temperatures with stock HSF are running 35-37C (CPU) and 70C (CPU Core). I'm sure why the core reads that high but the system seems to be pretty stable and fast! Has anyone been able to use the "yellow" RAID IDE connector for regular IDE hard drives? I did notice somehting interesting. The SilentTek monitor will completely shut off the CPU fan if it is enabled. Seems a bit counter productive.

JB

 

[KoolHonda]

"Has anyone been able to use the "yellow" RAID IDE connector for regular IDE hard drives?" Yeah, works just fine for me.

 

[Archangel07]

Cool another Geil Dragon user

Update, I have backed off the 252 to a 250, as I have decided to back off my voltage to 1.6. Before I stated that I thought I could get up to 255, but I have since reevaluated and found that 252 is the limit for my board with the present cooling I have. No matter what I did, I could not get the damn thing to run stable at 253. Anyway thought you may want to know.

Now for the mem. Have you tried ocing the Geil Dragon mem at 1.66 ratio, and if so what were your results. I attempted to oc at 250 with 1.66, which put me at 400mhz, well within specs, with no luck. The cpu is definitley not the prob, as this batch is highly ocable accoridng to the info, and the 1.33 ocability bears this out.

Finally, have you experienced any cold boot probs with 1.04a as others have.

 

[jbur293]

I have tried the Geil Golden Dragon 3200 at 1.66 at 250FSB and it worked fine in WinXp but would not complete the Prime95 torture test. At 1.66 the memory defaults to 2-4-4-7 but it works at 2-3-3-6 but no lower. The board doesn't seem to like to have the memory voltage at anything higher than 2.6v which is unusal since the memory spec is up to 3.1v. On another motherboard (Abit IS7), I have had the memory running at 1:1 up to 219Mhz without ptoblems but haven't used the 2:2 ratio on the AX4C Max. The Geil Golden Dragon is the best memory I have used to date. I have used OCZ and Kingmax and they couldn't overclock as well. For all of the fan fare for the Abit IS7, mine would not go above 250FSB regardless of ratio (5/4 or 3/2). The AX4C Max can go at least to FSB 260 but I haven't tried to push it higher yet. My current memory setting is 1.33@250 at 2.6v. The memory defaults to 2-3-3-6 at 1 clock at 1.33. I can run Prime95 all night with that setting. Unfortunately that's not to the memory's spec of DDR400.

I have had some problems with the cold boot with the 1.04a bios. That version did perform better than the 1.04 version though (minus the checksum error on cold boots).

Does the "yellow" IDE connector perform any better than the other two? I have two ATA100 drives on one of the "blue" connector but they wouldn't boot on the "yellow".

One last thing. On CPUz I noticed that PAT seems to be on regardless of the overclock unlike the Abit IS7. AX4C doesnt show the PAT notice on bootup but it appears to be on when in WinXP using CPUz. On the Abit, it would only work when the memory was at a 1:1 ratio.

JB

 

[duncan888]

There is a new bios out for this board now. I've just updated mine and will be doing some OC testing.

 

[CruisinGT]

Originally posted by: jbur293
Does the "yellow" IDE connector perform any better than the other two? I have two ATA100 drives on one of the "blue" connector but they wouldn't boot on the "yellow".

Are the 2 ATA100 drives similar (i.e. same capacity)? The yellow IDE connector is there if you want to setup your HD's in RAID. Both HD's must be the same capacity and is recommended that the HDs be the same brand/model although that is not a requirement. Once both are set in RAID, you should see an increase in data read/write throughput to your HD. It will be using both HD's together in parallel (assuming u set it in RAID 0) similar to how your chipset uses your RAM in dual channel right now. But just a warning about RAID 0, if one HD goes out than most likely all the information on the other HD will be inaccessible also. Considering the not so many times I had an HD completely crash on me, I took the risk. Good luck!

 

[Archangel07]

Thanks for the info on the Geil. I think I am having problems with respect to W2K. I had to disable HT as W2K does not recognize. I am thinking that the issue is with W2K, POST is fine, and I can sometimes get the OS to boot, but nothing close to stable. I think it is time I upgraded to XP. It has performed rock solid since getting up and running. I like to hear that a Abit user finds the Ax4C to be better. I am of the opinion that the Abit is a little over sold and the AX4C is under appreciated. One board will inevitably vary from the next, even from the same manufacturer. Abit just has a reputation for being an overclockers board.

New BIOS, that rocks. Be sure to post your results.

What happened to Shoman?

 

[Archangel07]

New BIOS is for DOS . . . you had me all excited for naught

 

[duncan888]

it means u can only install it from dos. So use a boot disc and update the bios that way!

 

[jbur293]

BIOS 1.07D update:

- Tried several different varieties of overclocks with not apparent change in results (1.6 and 1.33 ratios). If anything, the higher overclocks were not working as well.
- The only benefits that I have found so far is the "cold boot" problem" seems to have been corrected and I was able to lower my vcore voltage .025v to 1.625v on my 2.4-->3.0GHz overclock. The effective voltage is now around 1.58v.
- There were also some minor changes to the BIOS menus

-On another note I was able to finally get the "yellow" IDE connector to work with my drives. I had to go into the RAID utility and select option "1" for auto configure. At first I thought that I had inadvertantly created a "striped" array and had wiped out my hard drive contents but everything turned out OK - so far.

 

[Archangel07]

(Smacks hand against forehead) Damn I am getting stupid in my old age, here I am thinking that "for DOS" meant for Win98 or Win95. This flashing from windows is going to take some getting used to.


My rig is down right now due to change outs in my setup, are there any changes in VDIMM option, such as can we now go higher than 2.8v?

 

[jbur293]

No. VDimm is still limited to 2.8v. In fact, my board doesn't like it when I sent my memory above 2.6v. When I set it to 2.8 my system fails to boot. The memory worked at 2.8v fine on an Abit IS7 so I don't think it's the memory.

 

[CruisinGT]

Hey guys,

I recently tried overclocking my 2.53MHz system to about 166fsb. It worked fine with a single 256Mb module up to 180fsb running continuous MemTest loops but when I insert 2 256Mb modules in dual channel and attempt to run MemTest at 160fsb, it will always fail on Test5 at about 90% on the first pass. Actually, it fails at any overclock. I've tried it with both bios 1.04 and 1.04a. I've also already tried both modules seperately on the black slots with the same results.

Some other info:
I'm using OCZ PC3700 Gold modules at DDR320-366 at C2.5-3-3-8
vcore: 1.62v
vdimm: 2.7v
vagp: 1.65v
Spread Spectrumff

Any other ideas??

 

[CruisinGT]

No other suggestions as to what I should try or may be doing wrong??

I was thinking of getting a couple TwinMOS/Winbond PC3200 256Mb modules to see how I fare with those since it is in the AOpen's recommended DRAM list and I have heard good reviews from them with this motherboard.

 

[Archangel07]

No suggestions here on your prob, as I am also having some issues with my mem. Have you used the mem in another dual channel setup? i.e. GNB board or Nforce? and are the mem sticks a dual channel set? Could be some timing issues with the sticks as I have read other complaints regaridng OCZ sticks and timing errors. If the sticks are somewhat off kilter it could cause them to fail mem test. I would try relaxing the timings to see if that helps. While there is a lot of hoopla about CAS/RAS timings in that mem that can achieve lower is better than mem that cannot, it really does not make for all that much real world difference. The mem sticks your using are Samsung sanded down and a heatsink slapped on.

My prob stems from an inability to get my mem to operate at 433 spec. I can relax the timings to 2.5-7-4-4 and get it to run at 1.66 at a 245 fsb, but this puts me at 392. Installed new bios, and did not fix the problem. Only bottleneck seems to be with the memory, whether its a prob with the chipset or just a bad set of mem I am not sure, but I intend on finding out.

I have not had any problems with running my mem at 2.8, and all of these probs with mem, along with a change in the configuration of the board via Max II leads me to believe that there may be some problem with the pipeline.

And your right this board works well with Twinmos, but I like my dragon mem

 

[CruisinGT]

Originally posted by: Archangel07
No suggestions here on your prob, as I am also having some issues with my mem. Have you used the mem in another dual channel setup? i.e. GNB board or Nforce?

No....I currently have no other dual channel motherboards although NForce2 is looking pretty good for a multimedia PC right about now.

Originally posted by: Archangel07
and are the mem sticks a dual channel set?

The OCZ Dual Golds were but the Twinmos I just got are not.

Originally posted by: Archangel07
Could be some timing issues with the sticks as I have read other complaints regaridng OCZ sticks and timing errors. If the sticks are somewhat off kilter it could cause them to fail mem test. I would try relaxing the timings to see if that helps.

Hmmm...I had already tried relaxing the timings as far as they would go and increased the voltage allowable by my mobo, they just can't play nice with each other. I did have a little better luck with my TwinMos though. I was able to o/c it to 160fsb and ran a whole bunch of stability tests to verify it. I could boot the system up at 166fsb but it was not completely stable in MemTest though (particularly Test#5).

Originally posted by: Archangel07
My prob stems from an inability to get my mem to operate at 433 spec. I can relax the timings to 2.5-7-4-4 and get it to run at 1.66 at a 245 fsb, but this puts me at 392. Installed new bios, and did not fix the problem. Only bottleneck seems to be with the memory, whether its a prob with the chipset or just a bad set of mem I am not sure, but I intend on finding out.

I agree, I think there is an issue with either the chipset or its memory controller. I've tried many different types of RAM and so far I have never had any luck with OCZ with this motherboard. Twinmos is ok with this motherboard. And Kingmax (believe it or not) has allowed me the best overclock. :/ By the looks of it, the slower the RAM the better the motherboard behaves. I have yet to try the 1.07d bios. I've heard its been allowing better memory badwidth numbers but I haven't heard anything on better memory compatibility nor allowing higher overclocks. So the only thing I can tell you is that I wouldn't bet that you have a bad set of memory modules. Most likely it is the memory interaction with the motherboard, unfortunately. Well, I know how far I can go with TwinMos and Bios 1.04. I'll have to find time to test it with Bios 1.07d this week......and the different RAM combinations. and see if there is a difference myself.

Originally posted by: Archangel07I have not had any problems with running my mem at 2.8, and all of these probs with mem, along with a change in the configuration of the board via Max II leads me to believe that there may be some problem with the pipeline.

Me too. I often kick myself for not waiting a little longer for the Abit version...<sigh> Oh well, it offers a little more of a challenge for me, I guess. :/

 

[Archangel07]

Yeah sorry about the hurried response and failure to see your specs on timing.

Abit boards seem tobe havign the same issues with corsair, kingston, et al. hitting ceilings of 250 fsb or lower in some cases, and not even being able to oc at all.

I am realy starting to lose confidence in the review sites, including this one. You cannot rely on any reviews to determine the stability of the board, mem, et al. Sad when puter geeks are swayed by politics. Sad indeed.


I think your right, in that the lower speed ram is more stable with the dual channel setup. I can see the compatibility issues arising in respect to the 865 chipset, but you would think that the 875 would be better quality.

I am starting to wonder if stability is achievable with this board past certain fsb limits, Shoman above reported great results, but who the hell knows who he/she is or how reliable. There is no way to get this board above a 420 1:1 in my opinion with any stability. I am running 37c at load, 28c idle sys at 26c or 27c load and the highest I have been able to oc is to 415 and Prime still cuts out at about 3.5 hours worth of run time. I am going to move on to the 1.66 setting and see how far I can get, but I am sure as hell disappointed in not being able to get a decent 433 out of my mem. Really screwed up part about it is that even when I am able to get a stable oc the damn board is giving me errors on reboot, which makes no sense whatsoever.

But hey if it wasn't a challenge it would not be any fun as you pointed out.

Keep me updated on your progress with the twinmos. I ordered a IS7-G just to do some comparisons. One of these boards will get RMAd back, depending on who wins the stability test I can take the restock fee, and since I am in the legal profession I have a feeling that I can get them to waive it

 

[Archangel07]

By the way, the new bios doesn't do shite! for stability