P4 2.8 Slz5 + Epox 4pca3+ [Archive]

aristidis

12-20-2003, 05:39 AM

UNDERvolt?? I think your board is overvolting like crazy! If you set 1.65V in BIOS, and it reads 1.77V under load, it must be overvolting.

The voltage reading in CPUZ is the AGP voltage and not the vcore.
If you sea the vcore voltage at USDM you are going to understand what i mean. With this chip i was able to post at 287fsb with 1.8vcore but of course it is not stable at all. I want to burn in with Prime for several hours and going to higher fsb's slowly hoping to get much better resaults.

Blue078

12-20-2003, 07:24 AM

That looks to be a good chip you have their :)

My 2.8C non MO would only hit 3.5Ghz on water. So I'd say you got a good one :toast:

How is that GEIL PC 42OO working out ??

Shane5578

12-20-2003, 07:34 AM

i'm starting to see alot of happy epox users. lol

aristidis

12-20-2003, 11:20 AM

How is that GEIL PC 42OO working out ??

I only can saw you a Sandra Score with an 2.6C on my old MAX3
300fsb with timings 7-4-4-2.5 They are grate to me...

aristidis

12-20-2003, 11:26 AM

i'm starting to see alot of happy epox users. lol

Antil now it looks a nice board, but i want to sea what that chip can do on the MAX3 that arrives from day to day, beqause my old one is dead and gone for RMA.

Grace

12-20-2003, 11:50 AM

The Vcore on the Epox 4PCA3+ is rock stable within 0.03v from where you've set it at, idle and under load. Forget what the Winbond monitor tells you.

i want to sea what that chip can do on the MAX3 that arrives from day to day, beqause my old one is dead and gone for RMA.

You'll get about 100-150MHz higher at the exact same settings.

Shane5578

12-20-2003, 12:52 PM

my mum is called grace too. lol

it will be interesting to compare epox and abit.

retrospooty

12-20-2003, 01:16 PM

Originally posted by enroscado
UNDERvolt?? I think your board is overvolting like crazy! If you set 1.65V in BIOS, and it reads 1.77V under load, it must be overvolting.

My .02

BTW: nice chip tou got there

Those programs to read voltages are useless... the only way to really measure voltage is with a voltmeter.

Measuring real voltage, the 4PC3A v2.0 undervolts about .025 so your 1.65 is really 1.625

For me, using all the same components my MAX 3 maxed out at 270mhz FSB 5:4 , my Epox gets up to 288 FSB 5:4 .

aristidis

12-21-2003, 12:37 AM

You'll get about 100-150MHz higher at the exact same settings.

I hope that Grace, until now i am prime stable 275fsb with 1.725vcore and going up.

it will be interesting to compare epox and abit.

With my 2.6 L321 Max3 was better and more stable at high Fsb's.
Let's sea what will happens with this chip.:D

OC_Newbee

12-21-2003, 12:53 AM

Originally posted by enroscado
UNDERvolt?? I think your board is overvolting like crazy! If you set 1.65V in BIOS, and it reads 1.77V under load, it must be overvolting.

My .02

BTW: nice chip tou got there

Yeah aristidis right, mine also read the voltage wrong in cpu-z. It's reading the voltage of from my vagp instead of the vcore.

The vcore to this board undervolt to about .025-.05 in my board, but the vdimm and temp is pretty much on the money.

aristidis

12-21-2003, 01:49 AM

Yes this board undervolts at vcore and in mine it's about 0.05, but the most important is that vcore it isn't stabe in full load.
And a new pic of 276fsb

TASOS

12-21-2003, 03:49 AM

Overclockability & Epox ?

xxmm...it's just us Grace told you. 100-150mhz less than a Asus or Abit.

I have tested both the Canterwood and Springdale Epox mobos.

retrospooty

12-21-2003, 09:29 AM

I just used the Epox "magic BIOS" utility and got the 12/16 BIOS. I was able to eek out a few more mhz with it.

It has some new memory tweaks as well. :D

OC_Newbee

12-21-2003, 05:23 PM

aristidis

12-22-2003, 03:02 AM

I can't find the new bios anywhere on there site. Do I have to use the flashing utility that came with the board to update to the new bios?

Yep you have to use the magic bios utilitie. Just flash it but dont test it yet.

aristidis

12-22-2003, 10:53 AM

I also have noticed that the mofsets of that board are extremely hot and after placing some ram heatsinks and an 80mm fun on them i was able to run at the same clocks but with less 0.025 vcore and more stable ..:)

NaHeMiA

12-22-2003, 03:47 PM

VEry nice overclock :)

Would like to test a chip like that under my prom mk2 :D

OC_Newbee

12-23-2003, 10:34 AM

Originally posted by retrospooty
I just used the Epox "magic BIOS" utility and got the 12/16 BIOS. I was able to eek out a few more mhz with it.

It has some new memory tweaks as well. :D

I cannot use the Magic Bios Utility yet, due to the fact that I just move to a new home and haven't hook up the phone line yet(going through the process). I can download the bios from here(at work), but I cannot install the magic utility onto my work computer(it not an epox machine). So, can anyone post a link or something that would lead me to the new bios. I'm so anxious to test out this new bios. Thanks in a million.

Alexandrus

12-23-2003, 12:03 PM

Yeap, I want the new BIOS asa well, but no mention of it on EPoX's site.

retrospooty

12-23-2003, 05:04 PM

There is a post in this thread with a link.
http://www.aoaforums.com/forum/showthread.php?t=22231

OC_Newbee

12-23-2003, 06:11 PM

Originally posted by retrospooty
There is a post in this thread with a link.
http://www.aoaforums.com/forum/showthread.php?t=22231

I'd try that zip file from leotonch, and it's telling me that the bin file is not an award bios. Lemme try it again and see what sup.

aristidis

12-24-2003, 01:19 AM

New bios tested, but for me isn't better than the old one, 2-3 lower fsb for stable settings. And this new memmory setting it's better to keep it Disable..And....MERY CHRISTMAS TO EVERYONE:)

Alexandrus

12-24-2003, 03:18 AM

I'll try it today and let you guys know how it works.

dioSRL

12-24-2003, 11:09 AM

it's up&running.. fast POST, fast ide detection. Dunno, new memory settings.. does not affect performance.. i kinda like this new bios.

Alexandrus

12-24-2003, 11:33 AM

Yeap, same here, I like it as well, hope it works well for OCing, haven't tried that.

PS. You hairy fat ugly guy :P

Dr.Demonic69

12-24-2003, 12:30 PM

May i ask why are you running your vagp 1.8v??????
Isn't 1.5v default enough since these power 3d cards pump power from the psu molex connector?

dioSRL

12-24-2003, 12:45 PM

because raising vagp also rise the nb voltage and aids for stability

Craig

12-24-2003, 01:20 PM

So, if the Abit board will give him 100-150mhz more, then he'll have a 4 gig system with home made water blocks!

How about some pics of the blocks? And what sort of pump and rad are you making use of?

Thanks! !

Great set up! !:toast: :slobber: :toast:

OC_Newbee

12-24-2003, 02:32 PM

The new bios work out great for me. Currently running at 288fsb 1.70v bios(1.65v usdm), memory at 5:4 2-2-2-5 @3.2v memory aggressive mode on max, max load temp are at 53C. While with the older bios I'm only prime stable at the speed in my sig.

Been priming with two instance of prime95 overnight and still going this morning when I check. Should be home about after midnight to double check on it, so far this new bios is great.

Later when I get home tonight, I'm going to try for higher. I'll post back with the result.

Alexandrus

12-24-2003, 03:00 PM

Originally posted by Craig
So, if the Abit board will give him 100-150mhz mor
Who says ABIT will give him 100-150MHz more ?

Grace

12-24-2003, 03:45 PM

Originally posted by dioSRL
because raising vagp also rise the nb voltage and aids for stability

VAGP and VDD are NOT joined on this motherboard. It will aid stability too little and that's because it helps the AGP bridge to remain functioning.

If you want real results you need to do a VDD mod.

Who says ABIT will give him 100-150MHz more ?

I for instance.

The Epox 865/875 overclock 100-200MHz less that other motherboards at the exact same settings.

dioSRL

12-24-2003, 03:57 PM

i think just an epox guy could bring some light to this matter :) but ofcourse they will never tell:banana: Until then, I luv my pca3+, even when she's running at stock speed.

OC_Newbee

12-24-2003, 05:03 PM

Originally posted by dioSRL
i think just an epox guy could bring some light to this matter :) but ofcourse they will never tell:banana: Until then, I luv my pca3+, even when she's running at stock speed.

I love this board also, where my previous IC7-G max II with the same setting and parts can only do 275fsb and memory at 5:4 2-3-3-7.

Grace

12-24-2003, 05:13 PM

Originally posted by OC_Newbee
I love this board also, where my previous IC7-G max II with the same setting and parts can only do 275fsb and memory at 5:4 2-3-3-7.

The IC7 would hit at least 295 at the *exact* same settings.

The IC7 doesn't have a Vdimm of 3.2v, does it? Blame your memory.

retrospooty

12-24-2003, 05:22 PM

Originally posted by Grace

The Epox 865/875 overclock 100-200MHz less that other motherboards at the exact same settings.

Not entirely true. For some people, in some circumstances maybe.

With all the same parts (except the mobo) my MAX3 maxed out at 270 5:4 this Epox 4pc3a+ is stable up to 288, thats a full 288mhz faster at stable operating speeds. Of course that is mainly due to the MAX3 and its issues at 5:4, but its not always faster.

Both mobo's can hit 300 at 3:2

Soulburner

12-24-2003, 05:27 PM

Grace

12-24-2003, 05:35 PM

Originally posted by retrospooty
Not entirely true. For some people, in some circumstances maybe.

With all the same parts (except the mobo) my MAX3 maxed out at 270 5:4 this Epox 4pc3a+ is stable up to 288, thats a full 288mhz faster at stable operating speeds. Of course that is mainly due to the MAX3 and its issues at 5:4, but its not always faster.

Both mobo's can hit 300 at 3:2

Always entirely true, tested tens of Epox motherboards myself and people familiar to me. Always the Epox overclocked 100-200MHz less than the rest I ever got it against.

Your MAX3 maxed out due to problems with the divider that is known to have. That is a design fault, not something you can do anything about and it is unfair to compare like that. Throw the Asus P4C800 in which doesn't have divider problems, if you want to go with the 5:4 divider. I'm not talking that the MOTHERBOARD will go 5MHz higher on FSB but the CPU will go 100-200MHz faster. If both motherboards can hit 300 with the 3:2 divider, you can't compare using that CPU since it will max both motherboards already. If they could work over 300FSB stable the Abit would gain 5-10MHz more. When I say that they will go 100-200MHz faster, when a CPU and only a CPU is maxed out at 3500MHz, at another motherboard it will go to 3600-3700MHz at the same settings, given that the rest of components can follow also!

OC_Newbee

12-24-2003, 05:39 PM

Originally posted by Grace
The IC7 would hit at least 295 at the *exact* same settings.

The IC7 doesn't have a Vdimm of 3.2v, does it? Blame your memory.

On my previous IC7-G, Even if I set my ratio to 3:2 on the memory, I still wouldn't get past 275fsb. And all my part I have right now is the same exact part I have in the IC7-G.

I'm not talking about all IC7-G, I'm just talking about the previous one that I have.

Dr.Demonic69

12-24-2003, 07:06 PM

Originally posted by Grace
VAGP and VDD are NOT joined on this motherboard. It will aid stability too little and that's because it helps the AGP bridge to remain functioning.

If you want real results you need to do a VDD mod.

I for instance.

The Epox 865/875 overclock 100-200MHz less that other motherboards at the exact same settings.

My whole system is overclocked except for the vagp so should i increase it to like 1.6v to make sure everything is 100% stable???
Also can rasing the vagp mess up the 9800 pro????

Grace

12-24-2003, 07:22 PM

Originally posted by Dr.Demonic69
My whole system is overclocked except for the vagp so should i increase it to like 1.6v to make sure everything is 100% stable???
Also can rasing the vagp mess up the 9800 pro????

If it is stable, why mess with it?

1.6Vagp shouldn't be dangerous for the 9800Pro, just don't overdo it.

On my previous IC7-G, Even if I set my ratio to 3:2 on the memory, I still wouldn't get past 275fsb. And all my part I have right now is the same exact part I have in the IC7-G.

I'm not talking about all IC7-G, I'm just talking about the previous one that I have.

You did disable the last 2 settings in the 'game accelerator', right?

If you did, your motherboard had another problem. Hitting a hard wall at 275 with it is not normal.

retrospooty

12-24-2003, 07:24 PM

Originally posted by Grace
Always entirely true, tested tens of Epox motherboards myself and people familiar to me. Always the Epox overclocked 100-200MHz less than the rest I ever got it against.

Your MAX3 maxed out due to problems with the divider that is known to have. That is a design fault, not something you can do anything about and it is unfair to compare like that. Throw the Asus P4C800 in which doesn't have divider problems, if you want to go with the 5:4 divider. I'm not talking that the MOTHERBOARD will go 5MHz higher on FSB but the CPU will go 100-200MHz faster. If both motherboards can hit 300 with the 3:2 divider, you can't compare using that CPU since it will max both motherboards already. If they could work over 300FSB stable the Abit would gain 5-10MHz more. When I say that they will go 100-200MHz faster, when a CPU and only a CPU is maxed out at 3500MHz, at another motherboard it will go to 3600-3700MHz at the same settings, given that the rest of components can follow also!

Was this testing done on 4pca3+ version 1, 1.2, 2.0 or 2.2 ?

Anyhow as I said Not entirely true. For some people, in some circumstances , maybe even most of the time, but not in all cases.

Grace

12-24-2003, 07:30 PM

Originally posted by retrospooty
Was this testing done on 4pca3+ version 1, 1.2, 2.0 or 2.2 ?

Anyhow as I said Not entirely true. For some people, in some circumstances , maybe even most of the time, but not in all cases.

On plenty of all revisions.

And it is in all cases. The motherboard will limit the CPU top speed by 100-200MHz due to design. Epox skipped some improvements to make the motherboard more reliable, successfully. When it can reach a certain speed, any other board will overclock higher given that the board itself and the rest of the components can follow.

You want to test it? Get a 3.2C, a prometeia, every revision of every motherboard and see any Epox 865/875 fall back 200MHz behind.
Edit : Brands like QDI, Asrock and the like don't count. :p:

retrospooty

12-24-2003, 08:49 PM

If I need to get a 3.2ghz and Prommie to test your theory then it already proves mine... Not in ALL cases .... Make sense ? :D

I have tested my CPU on a 4PC3a+ IC7, IC7-G and a IC7MAX3 and in all 4 I can reach 3600mhz, and no more, that is my CPU's limit with water cooling, its the same with the Epox and the Abit's.

Now for me, with my BH-5 XMS 3500, running at 5:4 2-2-2-5 is the fastest overall performance. Given that fact my maximun stable 24/7 operating speeds are...

IC7 maxed out at 275mhz
IG7-G ............... 280mhz
MAX3 ............... 270mhz
4PC3A+ .............. 288mhz

Well, I gotta go do the family thing...

Merry Christmas :toast:

Grace

12-24-2003, 09:33 PM

Originally posted by retrospooty
If I need to get a 3.2ghz and Prommie to test your theory then it already proves mine... Not in ALL cases .... Make sense ? :D

Merry Christmas :toast:

Not any real sense at all.
If you count cases that the clock generator of the motherboard isn't to blame then it isn't the motherboards fault, so it is not even a case, it's merely a coincidence. The 3.2C cooled by a prometeia negates exactly that aspect. No hindering by the dividers or high FSB, only the CPU going to the top limit. You used 3 Abit boards with known 5:4 divider problems, hence they stopped not because of the motherboard being unable to push the CPU more but because the Abit motherboards gave up.

Edit : And actually the prometeia was a very bad example. Even the stock cooler, it won't matter, it will still point out the difference of the top speed. Under the exact same circuimstances, it does lose to the rest.

When the motherboards CAN get the CPU to run to the edge and aren't hindered by other factors, the Epox will lose by 100-200MHz, always. With an Asus P4C800 since you wanted to use the 5:4 divider, you would probably work fine up to 300FSB at the same settings. If you get to count cases that will include everything inside, they will be thousands of theories where one motherboard can be better than another. Even the worst of motherboards can be better than the rest in some such theoretical cases. Now if your Abit motherboards at 3:2 and high timings go as far as your Epox again at 3:2 and high timings, either you have a VERY special Epox or something else were wrong.

I recently switched from an Epox 4PCA2+ to an Asus P4C800-E. Running a 2.8C @ 1.8v exactly using the 5:4 divider, I had a top of 284FSB with the Epox and about 300FSB with the Asus, while I'm still going to work on it a little more when I get my watercooling setup running again after the Christmas. That's more than 200MHz difference, especially when comparing the Asus that under load the voltage drops severely.

Merry Christmas, have fun. :toast:

aristidis

12-24-2003, 11:48 PM

I’m not sure about this new bios settings but especially if you enable CPU parking the CPU really parks.
And another think. When MAX3 arrives I will have the opportunity to compare the two boards. But just for history with my previous CPU that was a 2.6c L321B134 on MAX3 was prime stable 295fsb and with EPOX 284fsb, always using memory 1:1.
So I hope that with MAX3 I will be able to hit 4Ghz.
I’m not saying that EPOX Is a bad board but it seams that it has a top limit.
:)

Grace

12-25-2003, 12:10 AM

Originally posted by aristidis
I’m not sure about this new bios settings but especially if you enable CPU parking the CPU really parks.

And another think. When MAX3 arrives I will have the opportunity to compare the two boards.

So I hope that with MAX3 I will be able to hit 4Ghz.

You should, given that we two have about the same cooling and I have a good M0 2.8C too...Yours looks like a nice piece as well. But mine is just...better (http://users.tellas.gr/~dawn/42.jpg).:D

If you can keep the water temperature around 20c, it should go to 4.1GHz with a very decent amount of voltage, not for 24/7 use though!

Have fun and Merry Christmas.:toast:

aristidis

12-25-2003, 01:05 AM

You should, given that we two have about the same cooling and I have a good M0 2.8C too...Yours looks like a nice piece as well. But mine is just...better.

:D I can borrow yours.....................Any way Merry Christmas.

And that is the best i can do prime stable with that board.

aristidis

12-26-2003, 01:58 AM

And a small update, after fixing a small problem with my watercooling system a was able to catch 280fsb and a max post at 4 ghz and it seams that the chip day byday became better:D

aristidis

12-26-2003, 02:00 AM

And my max post for the time.......................

Craig

12-26-2003, 01:24 PM

OK,

I get the fact that the Abit Max 3 and other Abit 875 versions have a flaw that really limits the performance of them.

And I will accept the position Grace posted regarding the Asus vs the Epox at the same settings with same components used.

This however leads me to this question.

Can the Asus, without a vmod, match or exceed the performance of the Epox if both are O/C'd to their limits with both boards stock?

Because I'm new enough to this that I'm seeking the very best O/Cing 875 board at stock, no soldering required. I have no problem with going a good bit farther for the cooling of the system, and motherboard, so long as I don't have to vmod anything.

Thanks

aristidis

12-26-2003, 01:50 PM

Because I'm new enough to this that I'm seeking the very best O/Cing 875 board at stock, no soldering required. I have no problem with going a good bit farther for the cooling of the system, and motherboard, so long as I don't have to vmod anything.

My opinion is that the best O/Cing 875 board without any modifications is the MAX3. I have tried even ASUS but you have to do soldering and cooling mods. The Epox is a fine board but it has a top O/Cing limit. I know that is a difficult choise and a bit of luck, but its up to you to decide.
:)

aristidis

12-26-2003, 02:15 PM

How about some pics of the blocks? And what sort of pump and rad are you making use of?

My Water cooling set up it has 3 blocks made from ELEVEN a Greek manufactore and specialy the Cpu block performs extremely good, for me even better than the Cascade, a Thermocill 120.1 rad. with 2X120mm ENERMAX funs, pump EHEIM 1250 with a small tank fit on and my litle secret, an extra water tank 300ml made from alu. All the the components of the water-cooling system are fited in the tower.
And a pic of the cpu block..

Grace

12-26-2003, 05:12 PM

Originally posted by Craig

This however leads me to this question.

Can the Asus, without a vmod, match or exceed the performance of the Epox if both are O/C'd to their limits with both boards stock?

That actually depends on the memory more than anything else. The only drawback of the Asus is the Vdimm limit that is 2.85v, not quite high for seasoned overclockers. If your RAM will work with that much of a voltage to the top speed that your motherboard/CPU will handle at the divider you want to run, then the Epox won't have any real advantage over the Asus, only drawbacks. However the Epox has a Vdimm up to 3.4v, the highest ever for a motherboard at stock. If you have voltage greedy memory, it is the one that can help it without voltage mods.

I would avoid the MAX3 as well. The divider and the VTT problems are a lot of headache. It would be the best if Abit would fix their own issues.

Soulburner

12-26-2003, 06:58 PM

And I wonder why the AI7 came out trouble free.

Its what the MAX3 was supposed to be for the 875.

Grace

12-26-2003, 07:04 PM

Originally posted by Soulburner
And I wonder why the AI7 came out trouble free.

Not completely, it still has issues with the dividers but at somewhat higher speeds. I would prefer it over any other of the IC7/IS7 line any day though.

Soulburner

12-26-2003, 07:06 PM

I haven't seen anyone here complain about divider problems with it.

However I was mainly pointing towards the whole VTT/vdimm thing they screwed up with the MAX3. They touted it as a big feature but then failed to deliver.

aristidis

12-27-2003, 12:42 AM

The MAX3 vdimm/VTT problem affects only vdimm reg. between 2.9V and 3.1V but if you use 3.2V this problem does not exists. As for deviders using memmory vdimm 3.2V there was no problem especially with the last bios and even the beta ones.

Grace

12-27-2003, 12:51 AM

Originally posted by aristidis
The MAX3 vdimm/VTT problem affects only vdimm reg. between 2.9V and 3.1V but if you use 3.2V this problem does not exists. As for deviders using memmory vdimm 3.2V there was no problem especially with the last bios and even the beta ones.

The VTT issue varies on each motherboard. Many motherboards will not work with 3.0v-3.2v as well while the 2.9v works. Willing to risk it?

Also the divider have nothing to do with the memory voltage. It cannot run by using the 5:4 divider at high FSB which differs with each board. True enough there are MAX3 motherboards that hit 300MHz with the 5:4 divider but one I've tested blacked out at 280FSB and I've seen reports of others failing even lower than that.

aristidis

12-27-2003, 04:13 AM

The VTT issue varies on each motherboard. Many motherboards will not work with 3.0v-3.2v as well while the 2.9v works. Willing to risk it?

I agree with you but you always take a risk buying a board and of course a chip or even memory. For examble i have tested two ASUS boards and both of them were bad pieces. As for the processors we all know what happens. Dont forget that you have test four SL6Z5 to find a good one.
As for MAX3 they said that for the new ones the VTT problem is fixed..