ASUS Maximus II Formula - new P45 king?

[radaja]

grnfinger&cryptik what phone# are u guys using to get a hold of someone at asus that can help in these situations.?

[CryptiK]

I didnt contact asus directly, I contacted the retailer I bought the board from and told them I wanted a new board and that they had to pretest it. Is there the number/contact email on your countries Asus page?

[Grnfinger]

grnfinger&cryptik what phone# are u guys using to get a hold of someone at asus that can help in these situations.?

These are the North American numbers and web site to trac your status which is a useless site.

http://livesupport.asus.com
510-739-3777 or 812-282-2787

[radaja]

grnfinger,thanks for all your help

[CryptiK]

Some 3D Benches at 24/7 settings with my old 8800GT:

Aquamark:


3D Mark 03:


3D Mark 05:


3D Mark 06:

[Rocks]

Cryptik,

I am trying desperatley to increase my RAM speed alongside my CPU speed. So far the highest speed I can get stable @ 4ghz is 942. I cannot run my RAM at 1255 no matter what I try. I can run it @ 1200 with a lower CPU speed but @ 4ghz I cannot get much bandwidth on my RAM. Which voltages do you suggest I play with the try to get stable? I've tried NB but to no avail. I'm thinking where I am at is most likely the best I can get. It is definitely solid but I want more!!

Thanks,

--Rocks

These are at idle:

Temperatures
Motherboard 34 °C (93 °F)
CPU 21 °C (70 °F)
CPU #1 / Core #1 36 °C (97 °F)
CPU #1 / Core #2 38 °C (100 °F)
CPU #1 / Core #3 34 °C (93 °F)
CPU #1 / Core #4 31 °C (88 °F)
North Bridge 42 °C (108 °F)
South Bridge 41 °C (106 °F)
PWM 33 °C (91 °F)
GPU 37 °C (99 °F)
GPU Memory 29 °C (84 °F)
GPU Ambient 30 °C (86 °F)
GPU VRM 31 °C (88 °F)

Cooling Fans
CPU 1480 RPM
GPU 990 RPM (62%)

Voltage Values

CPU Core 1.352 V
+3.3 V 3.264 V
+5 V 4.944 V
+12 V 11.928 V
+5 V Standby 5.139 V
CPU VTT 1.422 V
CPU PLL 1.574 V
North Bridge +1.1 V 1.422 V
South Bridge +1.1 V 1.177 V
South Bridge +1.5 V 1.501 V
DIMM 2.289 V
DIMM VTT 1.144 V
GPU Vcc 3.218 V

[CryptiK]

The thing is with 4GB of ram, getting over 1200 stable will be pretty difficult. I assume you are using 473 FSB x 8.5 = 4020.5 MHz?

If so, try the 5:6 divider, putting the ram speed at 1135 MHz.

Set:
Primary Ram Timings = 5-5-5-15-3-40-6-3 (leave the rest on auto)
Common Performance Level (PL) = 8
Static read Control = enabled
Dram Read Training = enabled
Ai Clock Twister = moderate
NB voltage = 1.45v
vdimm = 2.2v

Can you get that to run stable? What kind of bandwidth do you get compared to where you are at now?

[Rocks]

Awesome man, bandwidth is much better. I assume that I should enable the OC charger as well? How about the pull-ins? Keep them disabled? I running stress tests right now...seems stable.

My FSB is 471 so Mem is @ 1130. I will post the results in a sec I will edit this post.


Edit:

Original bandwidth:



New bandwidth:



I think its better, further tweaks could improve the write maybe?


--Rocks

[CryptiK]

That's much better. I'm not sure if you'll need OC Charger enabled, from what I can tell it may increase drive strength on the ram, and if you enable it you may be able to drop a bit off your vdimm. I run it disabled unless I'm clocking ram over 1200 MHz, in fact the ram will not go above 1210 MHz without it enabled. You may be able to run 2.0 - 2.1v vdimm since you're at 1130 MHz, we can try it first with mem OC charger disabled and see if your stable.

Ok now set:

Ai Clock Twister = 'Stronger', this will give you a drop in the latency and a small boost in bandwidth, if you can't stabilize it, try running it on 'strong'.
Sadly the Pull-ins do nothing as they do not work, so just leave them disabled.
Mem OC Charger disabled

If this is stable, try dropping your vdimm to 2.1v and re-test it.

[Rocks]

PC stuck on DRAM without OC charger enabled. If I enable OC charged still @ 2.2v DIMM, this is my bandwidth and just a fast test of prime 95 gave an error after 6 minutes.

This is with twister on stronger and DIMMv @ 2.2v.



I tried to run the twister on strong and stronger @ both 2.1 and 2.2 volts and neither would post without OC charger enabled.

--Rocks

[CryptiK]

Ok looks like your ram needs it then. Since you're getting an error and you're already at 1.45 vNB, try setting AI Clock Twister on 'strong' and see if your stable. If not, drop to 'moderate' and see if that helps.

[Rocks]

Strange I dropped the DIMMv to 2.1 and it seems prime 95 stable. Intelburntest bluescreened. You know, I am not so sure how far I should put my trust in these programs. I have yet to get a blue screen or a freeze with my overclocks on this board without running these stress testers.

I only get blue screens when I run these tests....go figure.

I'll up it to 2.2 v again and test it on strong I guess.

[CryptiK]

Since you bluescreened in linpack, might need to bump up your vcore, try 1.38v.

What are your four GTL Ref's set to?

[Rocks]

Ok I've raised the CPU core to 1.38, haven't tested it yet. My GTLs are:

GTL0 = +20mv
GTL1 = +40mv
GTL2 = +20mv
GTL3 = +40mv
NBGTL + +30mv

The 1.38 Vcore stablized it. My temps were through the roof though high 80s. Got a water cooler on the way, should be here tomorrow. My idle temps are solid as well as prime 95 temps. Prime 95 temps are mid 60s. Idle temps are mid 30 to low 40s. Thats is with a Tjunction of 100C. The water cooler I got is an all in one job. I'll link it up, but its in Japense so you might not be able to read it. It is the newer version of the EVERCOOL EC-WC-202.

It is the http://www.taoenter.co.jp/product/suireikit/WC-212.html

I am hoping it will cool better than my Silent square. The fact that it is water should help but I am a bit worried...

--Rocks

[CryptiK]

Try it as is, just increase vcore to 1.38v and run linpack 6 passes stress level 1.

If thats better, then try GTL Tweaking.

Set your GTL's to:

GTL 0 +10mv
GTL 1 -35mv
GTL 2 +10mv
GTL 3 -35mv

Does that make it better or worse?

[Rocks]

Stable 6 times no change in temps. I have loadline enabled so my vcore will hit 1.39 during load. temps hit the 90s with these settings. Hopefully the water cooler will bring them down.

[CryptiK]

Good to hear that stabilized it. I used to have a Silent square and even lapped my e6300 was pushing 76*C in prime. even a TRUE would bring them down 10 - 20*C, you should be mid to high 70's with that voltage. Hope the watercooling helps, I can't read japanese but that appears to be a small kit, have you read any reviews on it? Petra's Tech Shop is a great place for all your watercooling needs if you want to try and get something more hardcore.

[Rocks]

Here is my 3D mark 2006 score with these settings. I will increase the fan speed and see if it helps...

[GeorgeVasil]

After 20h searching for 9*500 stabillity..i want to share my result with u guys..

Lowest Vcore for Realtime 32M (no tweaks)



A litle bit vcore and we have wprime,vantage,pcmark etc..



Still searching for stability..i want to keep the voltage low..because a 0.04v up gives me +5 C under prime.

I tried several settings..

I tried the CPU Clock skew range from normal to 800ps..(it acts like not working)..

I tried GTL from .60-.67 ..nothing..and combitation of them (0.61 - 0.63 etc)..

My chipset temps are 47 idle (with water and new swifteck block) load under small ffts goes 52 and under blend goes at 58-60 with 1.33 vnb.I ensured the good contact twice.(crappy temps).

I think that my M2F get unstable because of high temps at chipset.

Rams@1000 2*1GB D9GMH for testing..l8r i will test with GSKill PI 4GB.

BIOS 1307

I want you opinion guyz..

[Rocks]

Good to hear that stabilized it. I used to have a Silent square and even lapped my e6300 was pushing 76*C in prime. even a TRUE would bring them down 10 - 20*C, you should be mid to high 70's with that voltage. Hope the watercooling helps, I can't read japanese but that appears to be a small kit, have you read any reviews on it? Petra's Tech Shop is a great place for all your watercooling needs if you want to try and get something more hardcore.

Prime 95 is mid to high 60s which is about the same I guess but when I use IBT, man it shoots up to high 80s low 90s. That program really puts out the heat.

[CryptiK]

@ GeorgeVasil - 4500 @ 1.288 is fantastic, nice chip you have there.

They are high chipset temps, especially for water. Have you placed washers behind the NB mounting screws? I just did that and the extra mounting pressure helped reduce the temps a few degrees. The chipset cooling design of these boards really leaves a lot to be desired.

When you use CPU clock skew, did you also use NB clock skew in conjunction with it? I found setting them the same (eg 300ps on both the cpu & NB) can improve stability in Orthos, but not in Linpack.

@ Rocks - yes linpack gets the cpu the hottest, but remember normal usage load temps wont ever approach those temps.

[Rocks]

@ GeorgeVasil - 4500 @ 1.288 is fantastic, nice chip you have there.

They are high chipset temps, especially for water. Have you placed washers behind the NB mounting screws? I just did that and the extra mounting pressure helped reduce the temps a few degrees. The chipset cooling design of these boards really leaves a lot to be desired.

When you use CPU clock skew, did you also use NB clock skew in conjunction with it? I found setting them the same (eg 300ps on both the cpu & NB) can improve stability in Orthos, but not in Linpack.

@ Rocks - yes linpack gets the cpu the hottest, but remember normal usage load temps wont ever approach those temps.

Yeah I agree, so what is the point in having it max out temps like that? It could give invalid BSODs.

If my water cooler can reduce my temps by at least 10c I will be happy.

[GeorgeVasil]

@ GeorgeVasil - 4500 @ 1.288 is fantastic, nice chip you have there.

They are high chipset temps, especially for water. Have you placed washers behind the NB mounting screws? I just did that and the extra mounting pressure helped reduce the temps a few degrees. The chipset cooling design of these boards really leaves a lot to be desired.

When you use CPU clock skew, did you also use NB clock skew in conjunction with it? I found setting them the same (eg 300ps on both the cpu & NB) can improve stability in Orthos, but not in Linpack.

No i Haven't placed any Washers..!I need a foto if you have.

Maybe the Probe reads wrong .. i wll measure it with a real sensor when i got one..

I will try to make a custom bottom plate for my block, I will make smth more rutty.

I will try the NB clock Skew...thanks for the advise.

ASUS should (no should ---> MUST) make a new bios in order to make all that settings work Properly.

Still Trying.. to get best Vcore for Prime..

[Eeky NoX]

Waoh! That's a great chip mate

So Cryptik if ya wana see some 3D Benchs, you should look at what I've done last night :

http://soj.mesdiscussions.net/soj/Ca...ujet_121_1.htm

I clam 3 walls yesterday, all on air : such a satisfaction my friends ya know ?


Sorry all cus' my thread is only in french, but any comment ??

[CryptiK]

71 470 marks 2003
29 258 marks 2005
19 640 marks 2006

Really nice results mate