*** Official ASUS P4C800/Deluxe (875P) Thread ***

[GNY]

Thanks Xeon I will probably go for the Best OCZ I can lay my hands on !!!

 

[Xeon]

Originally posted by: GNY
Thanks Xeon I will probably go for the Best OCZ I can lay my hands on !!!

One thing you might want to keep in mind when looking at their RAM is I've heard that the Modules they have that do the 3-4-4-8 timings are better suited for those with the Higher Speed cpu's such as your 3.0, so you might want to go that route rather than the sticks that do the lower latencies. Now I can't say one way or the other if that's going to make a big difference in your case since I don't have a 3.0 Ghz cpu, but you may want to send the guys at OCZ an e-mail and ask that question to be sure which will perform better for you.

Just a thought.

At any rate, keep us posted on your machines benchmarks once you get that baby built!

 

[Xeon]

Now that I've solved the MemTest-86/Asus P4C800-E Deluxe hang issue, I've spent most of today trying to find this boxes Highest FSB.

After hours of reboots and Bios Setting changes here is it's Highest Stable Limit is 279 FSB @ 1:1.

It's really odd though because I can boot into windows at a FSB of 288 no problem and even do normal work at this FSB. I think I'm hitting a wall on the actual cpu rather than a wall with the RAM itself.

I've only tried pumping the cpu vcore to 1.65 to see if it would make a difference, but it didn't, so I'm not sure what to think of those results from a cpu Vs. RAM limit standpoint.

At any rate if any of you have suggestions on what else I might try to get this FSB higher I'd love to hear from you.

Considering this RAM is PC-4000 Gold, it's flat out outperforming anything I'd expect it to do, so another big Thumbs up to OCZ for making such phenominal RAM. As I mentioned in a previous post those guys sent me several sticks to test to try and figure out this MemTest-86 issue and gave me pairs that went much faster than my original set. Their support has been nothing short of FANTASTIC!

 

[GNY]

Thanks for the heads up Xeon, I will call them on Monday and give them my specs and see what they recommend. 1 other thing Xeon, I"m thinking about buying the Heatsink you also recommended, What is the best Thermol compond to use on this puppy.

 

[Xeon]

Originally posted by: GNY
Thanks for the heads up Xeon, I will call them on Monday and give them my specs and see what they recommend. 1 other thing Xeon, I"m thinking about buying the Heatsink you also recommended, What is the best Thermol compond to use on this puppy.

The Thermalright comes with some so you won't need to use anything different right off the bat. As for which is the best, I really couldn't tell you but I do know that computer has some knowledge here so he may be able to comment further.

Arctic Silver 5 is probably a safe bet.

 

[MrSmithers]

GNY, I have heard that Shin Etsu makes the best stuff. However, the new artic silver 5 came out pretty recently, and I know there is a thread here on AT where a guy tried and really liked it, he had some before and after temps, but nothing scientific at all. I haven't seen any major places review the stuff though, so who knows. I myself have some of the artic ceramique stuff, and it seems to be working all right for me.

Smithers

 

[Xeon]

Originally posted by: GNY
Thanks for the heads up Xeon, I will call them on Monday and give them my specs and see what they recommend. 1 other thing Xeon, I"m thinking about buying the Heatsink you also recommended, What is the best Thermol compond to use on this puppy.

Here's an interesting article on the subject:


Thermal Paste Tests
Arctic Silver 5 Review1

 

[computer]

So Michael, we are still waiting on what fixed the Memtest problems! What was it, what did you do to fix them?

I got rid of them on mine by DISABLING the Promise controller. With in enabled, over 25,000 (TWENTY FIVE THOUSAND) error on the M$ RAM diagnostics and a bunch on Memtest. With it disabled, NO errors on the M$ test on the first pass and maybe one or two there after. Same with Memtest.

GNY, Nanotherm PCM+ is currently the best. See my post on it on this thread. Do a "find" on ....PCM+......not on this page or the last, but starting from the 3rd page back then go back. It's there somewhere.

 

[computer]

Ok I found one of my posts on Nanotherm PCM+.
JP, somewhere in this thread, or maybe another thread, I posted thermal compound results. All the websites I found showed Nanotherm PCM+ to be the best AVAILABLE, better than any Arctic product. I say "available" since there were two no-name beta products that where slightly better. (The beta items were from some new company that was "un-named"). I'm a reseller for Nanotherm as of a few days ago (since I found out about how good they were) and Scott told me they have a new product about to be released that's even better than the PCM+. I got my PCM+ about a month+ ago from http://ithinkpc.com/ for $9.95 free shipping. Their site is screwed up. You have to register just to browse find the PCM+! Here's the direct link. http://ithinkpc.com/info.cfm?productID=2686

 

[computer]

(DAMN SCREWED UP WEBSITE!!)

 

[computer]

(DAMN SCREWED UP WEBSITE!!)

 

[computer]

(DAMN SCREWED UP WEBSITE!!)

 

[Xeon]

Originally posted by: computer
So Michael, we are still waiting on what fixed the Memtest problems! What was it, what did you do to fix them?

I got rid of them on mine by DISABLING the Promise controller. With in enabled, over 25,000 (TWENTY FIVE THOUSAND) error on the M$ RAM diagnostics and a bunch on Memtest. With it disabled, NO errors on the M$ test on the first pass and maybe one or two there after. Same with Memtest.

GNY, Nanotherm PCM+ is currently the best. See my post on it on this thread. Do a "find" on ....PCM+......not on this page or the last, but starting from the 3rd page back then go back. It's there somewhere.

Actually, I would highly recommend that you look at the second link in my post above. It shows some pretty interesting facts about Nanotherm that you should be aware of.

As for the MemTest issue and how I solved it, I can't reveal that at the moment, you'll understand why real soon, but I will say, It was a much different problem than the one you were having.

 

[Xeon]

Originally posted by: computer
(DAMN SCREWED UP WEBSITE!!)

LOL, I'll assume you were having the same problem I've been seeing here on this site when you go to post, it takes you do a dead page and you don't realize until after you've hit the back button and tried to post again that your post has already posted.

I hate it when that happens. 🙂

 

[computer]

As for the Memtest issue and how I solved it, I can't reveal that at the moment, you'll understand why real soon, but I will say, It was a much different problem than the one you were having.

?????????????????????????????????? 😕 :disgust: :frown:

Yeah that's exactly what happened. This site is down every single day. You'd think they could get a server that works! It was down right before I was trying to post and every time I clicked the button to post it I got 'the page cannot be displayed". Years ago I got into the habit of right clicking my highlighted text and clicking "copy" for not if, but WHEN the site 'deletes' it, then I can just paste it and post again.

I forgot to mention about the memory; heck, OCZ is best out there. (Depending on what the heck was causing your problem) I'm more inclined to think some of us are just getting crappy P4C800xxxx mobo's and bad memory modules. Every big site you go to that tests hardware uses the P4C800xxx as their testbed, so we all KNOW it's the best 875/865 mobo out there. Also, when sites test other things they are more and more using OCZ instead of Corsair for their memory. I've seen many reviews and tests of the combo of OCZ and P4C800xxxx and have never seen anything but the highest of praise and awards for the combo. I think the "quality control" of these new PC hardware goods just plain SUCKS. Like M$ with their OS's, they are forcing them out on the market a year too early, and the QC suffers.

 

[computer]

Oh, and on the thermal compound review, I haven't read it yet but I will. Arctic Silver 5 is new and was not out when I read all the tests I found. Nanotherm however, does have a brand new "formulation" they told me, not yet released that is better than PCM+. I'm STILL awaiting the arrival of it from Scott @nanotherm/ESGN.

 

[shleepy]

Originally posted by: jhites

Originally posted by: CannibalisticH0b0
Timings: 2-2-2-5

vdimm, tried a few, but none made a difference.

Set the vdimm to 2.85v and leave it there. Set the Performance Mode under the Jumper Free Config to Standard. Set the vcore on the cpu to 1.5875v and reboot using the MemTest floppy to check stability after above settings. If you get errors, go into the chipset settings and change the Ras to Cas Delay to 3. Reboot to Memtest and check it again. You will have to play with the fsb settings and the memory timings to find your sweet spot. All cpu's and memory modules are different, so it may take you a while to achieve your maximum stable settings.

Hmm... well this sure is strange. I finally got around to doing this, but except for 2-2-2-5 @ 200MHz with PAT enabled, they all froze either @ 56% or 98%... Then, I did 2-2-3-5 @ 217MHz with PAT enabled, and that got like 135 errors, and then froze, but to my surprise, Windows actually managed to boot with this. Also, the 3DMark03 score improved a bit! I dunno whether I should keep this, go back to 2-2-2-5 PAT 200MHz (since that's nice and stable and has no errors with MemTest and doesn't freeze while running the program), or keep trying different timings... Suggestions? The timings that have frozen in Memtest or just don't boot up in Windows without the comp restarting right away are:

2-2-2-5 w/o PAT @ 217MHz
2-2-2-5 w/ PAT @ 217MHz
2-2-3-5 w/o PAT @ 217MHz
2-3-3-7 w/o PAT @ 217MHz

 

[sillious]

Originally posted by: jhites

Originally posted by: Xeon

PCMark2002 Scores

CPU score: 7630
Memory score: 8641
HDD score: 1099

If you have some scores please post them.

I did some a while back when I had first put this
system together that is in my sig.
CPU Score . . . . . 7882 . . . 8702
Memory Score . 10462 . . 10142
HDD Score . . . . . 1146 . . . 1191
The first set was running 3110Mhz @ DDR490 1:1 ratio and
the second set was running 3430Mhz @ DDR432 5:4 ratio.

FutureMark Link

I used PCMark 2002 to benchmark my system, standard vs overclock. The numbers in Bold are 20% oced.
My question is: the mobo is 800MHz FSB, but the FSB showing here is way lower. It is correct, if so, how can I raise the FSB? And could someone also say by the result, can I go any higher based on my configuration? I can't set it to 30% oc, it crashes. Another thing is the system has DX 9.1a, but it's showing 8.1, is it because PCMark2002 is older, or is it norm?

ASUSTeK Computer Inc.
P4C800-E
OS :Microsoft Windows XP
DX :8.1
CPU : Intel Pentium 4 2730 MHz
FSB : 200 MHz, 240 MHZ
Mem : 1022 MB


CPU Score = 6839, 8192
Mem Score = 8000, 8746
HDD Score = 1373, 1367

Thanks.

 

[MrSmithers]

The numbers you have there are correct. The stock bus speed is 200, and the P4 has always used a "quad pumped" FSB, for 800MHz, whereas the memory is DDR (double data rate) for 400MHz.

Smithers

 

[jhites]

Oriiginally posted by: sillious
My question is: the mobo is 800MHz FSB, but the FSB showing here is way lower. It is correct, if so, how can I raise the FSB? And could someone also say by the result, can I go any higher based on my configuration? I can't set it to 30% oc, it crashes. Another thing is the system has DX 9.1a, but it's showing 8.1, is it because PCMark2002 is older, or is it norm?

Take it off the AI Overclock and use manual settings for cpu and memory. You will most likely get better results and better overall performance.

 

[MrSmithers]

Xeon any ideas on when you will be able to reveal this magical solution to the rest of us. Approximate timeframe would be great if you don't know exactly.

Smithers

 

[shoman94]

Yea...might be a good idea...

One thing I did notice though is the board overvolts the vcore up to 1.600 then at 1.625 it under volts..... 1.650 and 1.600 in bios are the same once booted into windows...load and Idle.....
My board is set at 1.585 in bios...in XP idle is 1.65 and full load brings it down as low as 1.555..... THIS SUCKS. How can I get a stable overclock with those fluctuations????

 

[jhites]

Originally posted by: shoman94
Yea...might be a good idea...

One thing I did notice though is the board overvolts the vcore up to 1.600 then at 1.625 it under volts..... 1.650 and 1.600 in bios are the same once booted into windows...load and Idle.....
My board is set at 1.585 in bios...in XP idle is 1.65 and full load brings it down as low as 1.555..... THIS SUCKS. How can I get a stable overclock with those fluctuations????

Wow! Huge difference. Mine with the BIOS set to 1.585v in bios will show up to 1.60 in XP idle but will drop to between 1.500-1.525 under full load. These boards do have a pretty wide fluctuation range for most ppl.

 

[shleepy]

Errr... come someone please help me with my last post?

 

[shoman94]

Originally posted by: CannibalisticH0b0

Originally posted by: jhites

Originally posted by: CannibalisticH0b0
Timings: 2-2-2-5

vdimm, tried a few, but none made a difference.

Set the vdimm to 2.85v and leave it there. Set the Performance Mode under the Jumper Free Config to Standard. Set the vcore on the cpu to 1.5875v and reboot using the MemTest floppy to check stability after above settings. If you get errors, go into the chipset settings and change the Ras to Cas Delay to 3. Reboot to Memtest and check it again. You will have to play with the fsb settings and the memory timings to find your sweet spot. All cpu's and memory modules are different, so it may take you a while to achieve your maximum stable settings.

Hmm... well this sure is strange. I finally got around to doing this, but except for 2-2-2-5 @ 200MHz with PAT enabled, they all froze either @ 56% or 98%... Then, I did 2-2-3-5 @ 217MHz with PAT enabled, and that got like 135 errors, and then froze, but to my surprise, Windows actually managed to boot with this. Also, the 3DMark03 score improved a bit! I dunno whether I should keep this, go back to 2-2-2-5 PAT 200MHz (since that's nice and stable and has no errors with MemTest and doesn't freeze while running the program), or keep trying different timings... Suggestions? The timings that have frozen in Memtest or just don't boot up in Windows without the comp restarting right away are:

2-2-2-5 w/o PAT @ 217MHz
2-2-2-5 w/ PAT @ 217MHz
2-2-3-5 w/o PAT @ 217MHz
2-3-3-7 w/o PAT @ 217MHz

Try 2.5-2-2-6 then 2.5-3-2-6