2.4c + G.skill PC4400

[mtb856]

What I had to do around a 280 FSB (running ram 1:1), was put a blower underneath the ram moving air up towards the top of the case..... the blower is from a server HS, so it's fairly quiet but creates a good amount of high pressure airflow.

Here's a picture of it.... I superglued a piece of plexi over the mouth of the blower to direct airflow through my memory better. So far my memory isn't even warm to the touch anymore. BTW, this is with OCZ PC4000 EL Gold R2's.

http://img.photobucket.com/albums/v168/mtb856/Computer Mods/glowinside1.jpg

 

[Navig]

I've got one of those super talent RAM cooling mousetraps, but replaced the 40mm fans with some nicer ones. Plus I've got mounting tabs to squeeze a nice 80mm fan should I want to. The RAM chips just need some more surface area, so I'm waiting on those sinks.

 

[Navig]

Well, looks like temps were an issue. With my 3.2e and my IC7Max3 (vTT modded) I was able to get FSB 215 2-2-2-5 stable with 3.0V. I've got those thermaltake stick on memory sinks lining the non-labeled sides of the RAM sticks + supertalent 2x40mm. As you may recall, my unmodded P4C800e-dlx (2.85V max) could do FSB 210 2-2-2-5.

Will this translate into better high speed performance? Well, I'll swap in my 2.4c and find out...tomorrow.

(You may wonder why I had taken out my 2.4c, well I put the 3.2e back in to play Half Life 2! Just coincidentally finished, only an okay ending there. BTW, HL2 running with 2gb of 2-2-2-5 RAM was pretty much indistinguishable from running 1gb of 2-2-2-5, go figure).

navig

 

[Navig]

Well, I think I've hit the wall.

2.4c IC7Max3 with the following mods: Madcat vTT fix, Ramsinks, active cooled RAM, Asus northbridge sink + 40mm fan, Thermalright xp120 + Enermax 120mm, dual power supplies (OCZ 470W + Allied 300W).

Max fully stable FSB 275.
2.5-4-4-8/2.9V/AAAAA/Bios18

I tried to stabilize FSB 280, but couldn't quite get it there.
2.5-4-4-8/3.0V/AAAAA/Bios15 passed 28 loops of test#5, then failed.

These settings all failed quickly:
2.5-4-4-8/3.0V/AAAAA/Bios18
2.5-4-4-8/3.0V/ANADD/Bios18
3.0-4-4-8/3.0V/AAADD/Bios18
2.5-4-4-8/3.0V/AAADD/Bios15
2.5-4-4-8/3.0V/ANAAA/Bios15
2.5-4-4-8/3.0V/TAAAA/Bios15
2.5-4-4-8/3.0V/AEAAA/Bios15
2.5-4-4-8/3.0V/AA7AA/Bios15
2.5-4-4-8/3.0V/AA10AA/Bios15
2.5-4-4-8/2.9V/AA10AA/Bios15
2.5-4-4-8/3.0V/AAADD/Bios15

Here are some interesting notes. Bios15 definitely trended to slightly more stable than Bios18. I had some suspicions about Bios18 and high front side buses, during my experiements with a 3.2 SL7B8 14x multiplier. Now I have more suspicions.

I cannot for the life of me explain why the IC7Max3s max FSB would trail my P4C800e-dlx by so much. If anyone has any other ideas, please let me know.

Otherwise, I'm going to try out with my other IC7Max3.

navig

 

[Clevor]

Well, I think I've hit the wall.

2.4c IC7Max3 with the following mods: Madcat vTT fix, Ramsinks, active cooled RAM, Asus northbridge sink + 40mm fan, Thermalright xp120 + Enermax 120mm, dual power supplies (OCZ 470W + Allied 300W).

Max fully stable FSB 275.
2.5-4-4-8/2.9V/AAAAA/Bios18

Navig, what you are getting sounds about right. But 2.5-4-4-7 doesn't cut it for P4 and can easily be done using Hynix PC4000. How high do you think you will get at 2.5-3-3-7 , which is the main selling point of LEs?

Switch to 2.5-3-3-7 on both the Asus and Abit, and try running all these:

Aquamark, 3DMark2001/2003/2005, IL-2 demo for 30', SS2nd Encounter demo for full loop, GunMetal demos (both), PCMark2004 Pro (all tests), Codecreatures, UT2003, X-Isle, SuperPI 16M. Tell us where you can pass EVERYTHING. This is the criteria I use for a valid ram overclock.

Forget the Prime stuff; it's easy to get the rig 2D stable and Memtest clean at high fsb (I've hit 290 at 2.5-3-3-6 with PQI TCCD), but 3D stuff will fail Pronto. It's exactly the same phenomena with running 5:4 at high fsb on the P4.

I'm banking you will get to 271 or so on boards with PAT, if even that. I will be doing more testing here as soon as I get my replacement LEs.

 

[Stratcat]

Well, I think I've hit the wall.

I cannot for the life of me explain why the IC7Max3s max FSB would trail my P4C800e-dlx by so much. If anyone has any other ideas, please let me know.

navig

Have you tried cranking up the Vagp so more voltage gets to the NB - Sometimes works on an Asus, and an Abit owner (forgot who) in these forums claimed Vagp relates to Vnb on Abit too.

My apologies if you've already done this, & I missed it.

Strat

 

[J☼E]

I have pc4400 LE and I can't get it to run 1:1 at 290 fsb What timmings would you reccomend for a 5:4 divider at 290 fsb? and what voltage? its at 2.85v right now

 

[flapperhead]

Well, I think I've hit the wall.

2.4c IC7Max3 with the following mods: Madcat vTT fix, Ramsinks, active cooled RAM, Asus northbridge sink + 40mm fan, Thermalright xp120 + Enermax 120mm, dual power supplies (OCZ 470W + Allied 300W).

Max fully stable FSB 275.
2.5-4-4-8/2.9V/AAAAA/Bios18

I tried to stabilize FSB 280, but couldn't quite get it there.
2.5-4-4-8/3.0V/AAAAA/Bios15 passed 28 loops of test#5, then failed.

These settings all failed quickly:
2.5-4-4-8/3.0V/AAAAA/Bios18
2.5-4-4-8/3.0V/ANADD/Bios18
3.0-4-4-8/3.0V/AAADD/Bios18
2.5-4-4-8/3.0V/AAADD/Bios15
2.5-4-4-8/3.0V/ANAAA/Bios15
2.5-4-4-8/3.0V/TAAAA/Bios15
2.5-4-4-8/3.0V/AEAAA/Bios15
2.5-4-4-8/3.0V/AA7AA/Bios15
2.5-4-4-8/3.0V/AA10AA/Bios15
2.5-4-4-8/2.9V/AA10AA/Bios15
2.5-4-4-8/3.0V/AAADD/Bios15

Here are some interesting notes. Bios15 definitely trended to slightly more stable than Bios18. I had some suspicions about Bios18 and high front side buses, during my experiements with a 3.2 SL7B8 14x multiplier. Now I have more suspicions.

I cannot for the life of me explain why the IC7Max3s max FSB would trail my P4C800e-dlx by so much. If anyone has any other ideas, please let me know.

Otherwise, I'm going to try out with my other IC7Max3.

navig

the asus is designed so that the nb voltage traks with the vcore voltage. so at i.6 volts vcore = 1.6 nb voltage. the abit ic7 series comes with a nonadjustable nb setting (approx)1.50 volts. now the max series may have a little more, but not enuff to challenge the ingenious design of the asus. i did the all in one mods(agp/nb/vdimm/vtt ) to my ic7 and it turned into an oveclocking monster.... when i hit the vagp with 1.8 volts and the nb with 1.9 v. it just took off,easily running to 340/350 mhz on a prommy

 

[Navig]

Joe, I'm not sure what timings would be best. I kindof skipped from 2-2-2-5 to 2.5-4-4-8 without testing limits in between. You can always just start with 2.5-4-4-8 and lower as you go.

flapperhead, since I've got 2 Max3s, I'm think I'm going to vagp/nb mod it. I just hate cutting traces. I do think this is the prob. Vagp is currently set to its max 1.65V. Did you use hipro's mods?

Kindof ironic that Asus will put more volts thru the northbridge, what with that passive sink and all.

Anywho, going to try it in my other Max3.

navig

 

[Navig]

Hey, what do you know? Double Prime95 stable at stock speeds at 1.434V under 100% load, not bad for a northwood. I love running this motherboard with woutertal's droop mod, I get to set the voltage to 3 decimal places to whatever I want. I wonder how low I could take it? I really need to droopmod my P4C800.

Okay onto some overclocking.