ASUS Maximus II Formula - new P45 king?

[CryptiK]

Just mucking around with it today:

500FSB, CellShock GKX @ 600 MHz CL 5


539 FSB, CellShock GKX @ 1078 CL 4


570 FSB (can't go any higher with this board although CPU does 590 FSB on same vcore on P5Q-Deluxe

[Cable_Tie_Guy]

3.5GHz
500FSB
DDR2-1200 5-5-5-18-2T

[d3885u]

Is there anyone of you can pass the FSB 600 please...???
I'm still waiting for my Maximus II Formula coming soon...hopefully tommorow I could testing my M2F...and when I read about the thread and sharing posting...any reason why Asus Maximus II Formula didn't match the expectation for a R.O.G series motherboard...???
Why Asus is a more looks like "abbandoned" the M2F and the other P45 chipset...altough on the other side...Biostar,DFI,and Gigabyte looks like they succeed to create "THE FSB MONSTER..."
about the "bug" and the other "disfunction" about max FSB wall and the max RAM spd...Is it because of the BIOS, (coz I think the BIOS from ASUS is still looks so "premature"...) or because the limitation from the motherboard it's self...???Or maybe because Asus had the "defective" chipset of P45...???Thx 4 your sharing and the explanation 4 me...

[CryptiK]

I have no idea why this board doesn't perform as well as the P5Q-Deluxe, nor do I understand why there has not been sufficient bios development to keep it competitive. You are correct in saying Gigabyte seem to have done a much better job than Asus regarding bios development for their P45 boards, with 600 - 650 FSB being easily achievable and also they have great ram overclocking ability.

The maximus II formula is absolutely terrible at ram overclocking, I cannot get it to do DDR 1220 CL 5 no matter what. This is equivalent to the max on the Asus P5Q-E, a low-end board compared to this so called 'high-end' ROG board. My P5Q-Deluxe benched the same kit of ram at 1333 and I could load windows at 1380 and POST at 1400 using only 2.3v!

I cannot explain it, this board is supposed to be high end, hence it's 'ROG' classification and massive price tag, yet my P5Q-Deluxe (and everyone else's - go look at the P5Q-Deluxe thread for more examples) is better at FSB overclocking and way better at ram overclocking. The M2F also takes more vNB to run a given FSB and PL (tRD) compared to the P5Q-Deluxe. on the M2F it took 1.37v vNB to load windows at 570 FSB and it was very unstable, whereas I could bench 3D etc at 572 FSB at only 1.26 vNB on my deluxe.

Unless they can make this board at least as good as the P5Q-Deluxe at both FSB (620+) and ram overclocking (1300 - 1400+) then this board should stayed away from. Sure it can do average clocks and average FSB, but it is no fun, and definitely not deserving of its 'high end' classification.

Needless to say, when my P5Q-Deluxe returns from RMA, I will be changing straight back to it and getting rid of the maximus II formula.

[turtletrax]

Seems every Asus board I go to I get a kick in the jaw. This is pathetic at best.


If I cannot get 4.5Ghz 24/7 out of my E8400 I will most likely sell this board and move on. TOo bad really, this board sure looks the part.

[Cable_Tie_Guy]

Maximus II Formula
1:1 (400FSB / DDR2-800)
PL8

So, is PL8 possible with 1:1 on MF2, or it is because of the low freq mem?

[kwalker]

Ai Overclock Tuner: Manual
OC From CPU Level Up: Auto
Ratio CMOS Setting: 8.5
FSB Frequency: 444
CPU Clock Skew: Auto 
NB Clock Skew: Auto
FSB Strap to North Bridge: Auto
DRAM Frequency: DDR2-943MHz

DRAM CLK Skew on Channel A1: Auto
DRAM CLK Skew on Channel A2: Auto
DRAM CLK Skew on Channel B1: Auto
DRAM CLK Skew on Channel B2: Auto
DRAM Timing Control: Manual

1st Information: 

CAS# Latency: 5 DRAM Clocks
DRAM RAS# to CAS# Delay: 5 DRAM Clocks
DRAM RAS# Precharge: 5 DRAM Clocks
DRAM RAS# Activate to Precharge: 15 DRAM Clocks
RAS# to RAS# Delay: Auto
Row Refresh Recycle Time: Auto
Write Recovery Time: Auto
Read to Precharge Time: Auto

2nd Information:

Read to Write Delay (S/D): Auto
Write to Read Delay (S): Auto
Write to Read Delay (D): Auto
Read to Read Delay (S): Auto
Read to Read Delay (D): Auto
Write to Write Delay (S): Auto
Write to Write Delay (D): Auto

3rd Information:

Write to PRE Delay: Auto
Read to PRE Delay: Auto
PRE to PRE Delay: Auto
All PRE to ACT Delay: Auto
All PRE to REF Delay: Auto
DRAM Static Read Control: Auto
DRAM Read Training: Auto
MEM. OC Charger: Enabled
Ai Clock Twister: Auto
Ai Transaction Booster: Auto


PCIE Frequency: 100

CPU Voltage: 1.3675
CPU PLL Voltage: 1.60600
FSB Termination Voltage: 1.35175
DRAM Voltage: 2.21075
North Bridge Voltage: 1.40475
South Bridge 1.5 Voltage: Auto
South Bridge 1.1 Voltage: Auto

CPU GTL Reference (0): Auto
CPU GTL Reference (1): Auto
CPU GTL Reference (2): Auto
CPU GTL Reference (3): Auto
NB GTL Reference: Auto
DDR2 ChA Reference Voltage: Auto
DDR2 ChB Reference Voltage: Auto
North Bridge DDR Reference: Auto 

Still a lot of auto going on here but this is the only settings I can get the ram to run 1066.
the bandwidth is better at 471 4GHz but The Dominators with other than Microns wont agree with this P45 chipset.
I broke my PC10000's
I forgot to mention here that I'm running Vista 64 bit and 4X1gb pc8500's ......that makes a difference on how the ram responds


This screen shows the ram but bsod at reboot so I raised the NB voltage.
Seems ram is my issue and not CPU at all.
I toyed with the ref voltages and just about every other settings you can think of but no go on the ram OC without a latency change.
I'll post an update with different ram later.

[CryptiK]

Is there anyone of you can pass the FSB 600 please...???

I'm pretty sure 600 FSB would take upwards of 1.6v vNB if it is even achievable on most M2F boards, unless it happens to luckily have a good clocking MCH. Even with 1.45v vNB my board is at its limit at 570 FSB, 572 FSB will not POST. I think it would take 1.45 - 1.5 vNB to stabilize 550 FSB in orthos/linpack. At 1.40v mine fails Orthos in 2 seconds at 550 FSB

One guy (Solarfall I think his name was) did hit 600FSB, but he did not provide details of what vNB etc this required, nor his cooling setup.

[grifonemoloch]

I'm pretty sure 600 FSB would take upwards of 1.6v vNB if it is even achievable on most M2F boards, unless it happens to luckily have a good clocking MCH. Even with 1.45v vNB my board is at its limit at 570 FSB, 572 FSB will not POST. I think it would take 1.45 - 1.5 vNB to stabilize 550 FSB in orthos/linpack. At 1.40v mine fails Orthos in 2 seconds at 550 FSB

One guy (Solarfall I think his name was) did hit 600FSB, but he did not provide details of what vNB etc this required, nor his cooling setup.

Hi all! My first post on this thread, so greetings
Another user managed to be stable at 570FSB with bios 603, this is the screen he linked in his post (the post is on page 20 of this thread, first post of the page)

http://spacemaster.sx-team.com/Testo...like/E8500.JPG

I'm going to put a Q9550 and some G. skill RAM soon on this board, i will let you know my results

Bye

[CryptiK]

Thanks for that, I've been trawling the thread and I also found a guy who got 555 FSB Orthos stable. There is, however, definitely a consensus that the board isn't capable of much more than 600 FSB (607 is the highest I've seen).

I have a theory on that though. The board is not able to clock ram to 610 MHz, as I have extensively tested this with kits capable of 675 MHz+ . It can do 600 MHz, but I cant get 605 orthos stable. This would mean that 610 FSB would not be stable either, and around 600 - 610 FSB would be the absolute limit of the board.

I am sure that the MCH is limiting me though, as if I try 570 FSB on 1.32v vNB, it wont POST, but if I increase vNB to 1.35v it wil post but isn't stable, and I gain some (very little) stability by increasing that further to 1.40v. Above 1.40v it gains me nothing. Loosening PL to 13 from 10 allows it to POST at 1.32 vNB but it wont load windows, however I should not have to go higher than PL 10 at less than 600 FSB. Furthermore I have confirmed this CPU does 590 FSB on a deluxe at the same vcore (1.33v).

EDIT - I also do not know where/how Solarfall got bios version 603, all i can find is 604. I have tried 1307, 701 & 801 and all of them had the same limits, but 1307 recovers from a failed overclock properly, whereas the older bioses do not, requiring a power off at the PSU switch.

[GaBBa-Gandalf]

Did u tried the modBIOS 1307?

[CryptiK]

Yes, I'm the guy that asked Ket to mod it for me and was the first to try it out It didn't help my max FSB though. I really hope there is a new 'improved' bios released soon.

[GaBBa-Gandalf]

could u tell me please which changes has this "new" bios?

[CryptiK]

You mean what Ket changed in the 1307 mod bios? He added additional cpu support, and replaced the M2F memory table with the P5Q-Premium memory table. He may have made other small changes too that I'm not aware of.

[GaBBa-Gandalf]

ok... sounds interesting.. maybe i should test this mod bios...
Did it works fine?

[CryptiK]

Yeah absolutely, Ket's good at what he does, no problems at all, and I was able to overclock the ram a little higher too.

[d3885u]

Wow..that's great...where I can get the BIOSmod version is?
And sorry maybe I have some question to you Cryptik...Is there any different when you use the RAM slot at 1&3 comparing with slot 2&4...???
Which is the best between two dual channel option is?thx...

[64dragon]

i just joined the M2F club I received my board today and am currently replacing the stock TIM. I think Asus has figured out that theres issues with the nb being hot, the back of my board has 3 of their washers on it at the 2 nb screws.

[unex|st]

I have no idea why this board doesn't perform as well as the P5Q-Deluxe, nor do I understand why there has not been sufficient bios development to keep it competitive. You are correct in saying Gigabyte seem to have done a much better job than Asus regarding bios development for their P45 boards, with 600 - 650 FSB being easily achievable and also they have great ram overclocking ability.

The maximus II formula is absolutely terrible at ram overclocking, I cannot get it to do DDR 1220 CL 5 no matter what. This is equivalent to the max on the Asus P5Q-E, a low-end board compared to this so called 'high-end' ROG board. My P5Q-Deluxe benched the same kit of ram at 1333 and I could load windows at 1380 and POST at 1400 using only 2.3v!

I cannot explain it, this board is supposed to be high end, hence it's 'ROG' classification and massive price tag, yet my P5Q-Deluxe (and everyone else's - go look at the P5Q-Deluxe thread for more examples) is better at FSB overclocking and way better at ram overclocking. The M2F also takes more vNB to run a given FSB and PL (tRD) compared to the P5Q-Deluxe. on the M2F it took 1.37v vNB to load windows at 570 FSB and it was very unstable, whereas I could bench 3D etc at 572 FSB at only 1.26 vNB on my deluxe.

Unless they can make this board at least as good as the P5Q-Deluxe at both FSB (620+) and ram overclocking (1300 - 1400+) then this board should stayed away from. Sure it can do average clocks and average FSB, but it is no fun, and definitely not deserving of its 'high end' classification.

Needless to say, when my P5Q-Deluxe returns from RMA, I will be changing straight back to it and getting rid of the maximus II formula.

I totally agree with you, I spent some nights trying to get the max out of my 2x2gb PC8500 Dominators and I found that MIIF couldn't go any higher than my old P5B deluxe. Stock frequencies needed less voltage, but this is not exactly what I'm expecting from this board.

I played for hours with the "advanced settings" but I always ended up leaving them on auto. The only improvement was by delaying 100ps on B1 slot.

I found the best divider to be 3:4, all others were less stable or unable to boot (even @ PL 10). I found also some fsb holes, I can do a 420MHz stable but 421 won't post, with 430MHz back stable again. The same was for 450MHz, with the system becoming bootable again at 460MHz.

The only setting which actually made me gain some fsb MHz was tWTW_S. Loosening it from 4 to 5 I got about 7/8MHz without almost no bandwidth loss. Clock twister settings were again a no go: no noticeable bandwidth gain from light to normal and no boot with strong/stronger (even at stock DRAM frequencies).

I didn't try pull-in settings, but I don't think they would make the situation look different..

[Cable_Tie_Guy]

I totally agree with you, I spent some nights trying to get the max out of my 2x2gb PC8500 Dominators and I found that MIIF couldn't go any higher than my old P5B deluxe. Stock frequencies needed less voltage, but this is not exactly what I'm expecting from this board.

I played for hours with the "advanced settings" but I always ended up leaving them on auto. The only improvement was by delaying 100ps on B1 slot.

I found the best divider to be 3:4, all others were less stable or unable to boot (even @ PL 10). I found also some fsb holes, I can do a 420MHz stable but 421 won't post, with 430MHz back stable again. The same was for 450MHz, with the system becoming bootable again at 460MHz.

The only setting which actually made me gain some fsb MHz was tWTW_S. Loosening it from 4 to 5 I got about 7/8MHz without almost no bandwidth loss. Clock twister settings were again a no go: no noticeable bandwidth gain from light to normal and no boot with strong/stronger (even at stock DRAM frequencies).

I didn't try pull-in settings, but I don't think they would make the situation look different..

That is strange... I am happy with my M2F...
I can post with strong / stronger settings on any FSB / MEM / Ratio / Clock.
I can do DDR2-1200 with any BIOS I try...
I can do PL8 with 1:1 divider too, and boot with 400 strap if needed.

M2F let me overclock my quad with less volts then any other board I know.
And it is 500FSB stable on my quad, and I am all aircooled LOL!

All I did was the washer mod, and applied AS5 on NB and SB.
Without crazy volts, my NB is always below 38c on load. With crazy volts like 1.5v, it is always under 42c under load.

I am stable with a 65nm QX6700 (2.66GHz / 266) B3 HEAT MONSTER aircooled at:

7x500 = 3.5GHz
8x475 = 3.8GHz
8x450 = 3.6GHz
9x400 = 3.6GHz

I can post on 4GHz but it won't even load windows without insane volts.

If the board is doing 3.8GHz with a 65nm B3 (LOL) I bet with a good 45nm quad, this board can do 4.0 - 4.2GHz stable on air, and with a good water system, 4.2 - 4.5GHz.

With a good 45nm dual, it should be possible to be stable aircooled @ 4.3GHz or more.

This is a great board.

500FSB with 65nm B3 Quad and DDR2-1200 CL5


3.6GHz 450FSB with B3 Quad aircooled stable 24/7


My rig


Lapped stuff:

[Grnfinger]

Yes, I'm the guy that asked Ket to mod it for me and was the first to try it out It didn't help my max FSB though. I really hope there is a new 'improved' bios released soon.

I found his bios did nothing for my setup, no FSB improvement and no bandwidth improvement. I was hoping for a tref adjustment to improve badwidth sadly no.

[CryptiK]

That is strange... I am happy with my M2F...
I can post with strong / stronger settings on any FSB / MEM / Ratio / Clock.
I can do DDR2-1200 with any BIOS I try...
I can do PL8 with 1:1 divider too, and boot with 400 strap if needed.

M2F let me overclock my quad with less volts then any other board I know.
And it is 500FSB stable on my quad, and I am all aircooled LOL!

All I did was the washer mod, and applied AS5 on NB and SB.
Without crazy volts, my NB is always below 38c on load. With crazy volts like 1.5v, it is always under 42c under load.

I am stable with a 65nm QX6700 (2.66GHz / 266) B3 HEAT MONSTER aircooled at:

7x500 = 3.5GHz
8x475 = 3.8GHz
8x450 = 3.6GHz
9x400 = 3.6GHz

I can post on 4GHz but it won't even load windows without insane volts.

If the board is doing 3.8GHz with a 65nm B3 (LOL) I bet with a good 45nm quad, this board can do 4.0 - 4.2GHz stable on air, and with a good water system, 4.2 - 4.5GHz.

With a good 45nm dual, it should be possible to be stable aircooled @ 4.3GHz or more.

This is a great board.

I can do 1200 on the ram easily too, I said 1210 was unstable and 1220+ was no go. Try 1220 - 1250 for me and see how you go. Other P45's can do 1250 - 1350 easily and the M2F just simply cannot.

Any P45 board will let a decent quad do ~500 FSB pretty much stable, it's nothing really, and 4.3 on a wolfdale is also a gimme with the right vcore.

I have tried both the P5Q-Deluxe and the M2F and I can categorically state the P5Q-Deluxe is way better. Guy with quads in the deluxe thread have also easily hit 500 FSB and have done 620 FSB+ on wolfdales with stock cooling on the NB with moderate (<1.32v) vNB.

Try this on your M2F, it simply isn't going to happen:

DDR1333 MHz dual channel 2.29v (real) @ 1.26v vNB:

[Cable_Tie_Guy]

CryptiK,

Why da_hell am I going to do that?
You have there a DDR2-1333 CL5 at 444FSB, and you have 9K read and 60ns latency. That is so bad it hurts.

450FSB / DDR2-900 CL4 1:1 will do about the same bandwidth and latency you are doing on your P5Q with insane DDR2-1333 CL5.

Get real, you won't have more then 1-3% boost on any game going from reasonable DDR2 clocks, to insane DDR2-1333 like that.

So if I can have the same performance using DDR2-900 CL4 with 1:1 450FSB, why am I going for DDR2-1333 on this stupid P45 with PL limitations?

If you wanna talk BIG about memory overclocking, then show me something like this:



This was done about 18 months ago LOL!
That was DDR2-900 CL3 linked / sync on 680i. Used only 2.4v on mem.
Now, for those latency and bandwidth I fought.
P45 (any board) is not that good on latency / bandwidth, it has PL limitations. So there is absolute no reason to go for those clocks on ANY P45 board since the results are always bad.

If your thing is memory overclocking, you are for sure using the wrong board, been that a M2F or a P5Q-D.

Now we can talk about fsb limitations. That is fine. But don´t show me some DDR2-1333 on 9K read and 60ns latency on a P5Q like it is a good thing. This is a very very very bad thing. One of the worst performances I have ever seen on a high clocked DDR2. On my 680i I would beat that with DDR2-800 LOL.

And my quad is a 65nm B3 LOL. It is almost impossible to overclock it on air. What I did on this board with it is not that easy as you are saying.
My 45nm will arrive soon, then we will see what it can do on M2F aircooled.

[CryptiK]

CryptiK,

Why da_hell am I going to do that?
You have there a DDR2-1333 CL5 at 444FSB, and you have 9K read and 60ns latency. That is so bad it hurts.

450FSB / DDR2-900 CL4 1:1 will do about the same bandwidth and latency you are doing on your P5Q with insane DDR2-1333 CL5.

Get real, you won't have more then 1-3% boost on any game going from reasonable DDR2 clocks, to insane DDR2-1333 like that.

So if I can have the same performance using DDR2-900 CL4 with 1:1 450FSB, why am I going for DDR2-1333 on this stupid P45 with PL limitations?

If you wanna talk BIG about memory overclocking, then show me something like this:

*pic removed*

This was done about 18 months ago LOL!
That was DDR2-900 CL3 linked / sync on 680i. Used only 2.4v on mem.
Now, for those latency and bandwidth I fought.
P45 (any board) is not that good on latency / bandwidth, it has PL limitations. So there is absolute no reason to go for those clocks on ANY P45 board since the results are always bad.

If your thing is memory overclocking, you are for sure using the wrong board, been that a M2F or a P5Q-D.

Now we can talk about fsb limitations. That is fine. But don´t show me some DDR2-1333 on 9K read and 60ns latency on a P5Q like it is a good thing. This is a very very very bad thing. One of the worst performances I have ever seen on a high clocked DDR2. On my 680i I would beat that with DDR2-800 LOL.

And my quad is a 65nm B3 LOL. It is almost impossible to overclock it on air. What I did on this board with it is not that easy as you are saying.
My 45nm will arrive soon, then we will see what it can do on M2F aircooled.

Like in my GTL+ Ref thread, you seem to be an expert in missing the point. I was saying the M2F has memory overclocking problems, and it cannot clock ram above 1210 MHz. I then gave an example of how high the P5Q-Deluxe can overclock ram. I wasn't 'talking big about memory overclocking'. I was comparing 2 similar boards ability to overclock ram in terms of speed achievable. It appears you also misunderstood what I did, thinking it was my '24/7' ram setting. It was a ram overclocking test, simply to establish what those sticks would do on less than 2.3v real.

You then belittle the result, saying you can beat the efficiency, bandwidth and latency using a lower ram speed & lower CL. When did I ever say that I was trying to achieve any of those things? I didn't, I said 'ram overclocking' which meant 'ram overclocking' and nothing more.

I am fully aware of the low efficiency and high latency of the P45 chipset compared to 680i nvidia platform and P35. You don't need to tell me something I already know.

Furthermore, you say I should be on a board that is absolutely renowned for killing D9 based ram to overclock my D9 based ram. I don't know what you're thinking suggesting that, as using a board that will kill my ram is not a smart choice by any stretch of the imagination.

In the ram overclocking threads, where the sole aim is speed at a given CL or voltage, do you see guys saying where is your efficiency? No, speed is the sole aim. That was my sole aim using the P5Q-Deluxe, and I achieved that with 2.29v real.

Let me give you some advice - If you want to use the know-it-all tone you consistently do, and 'talk big', perhaps wait until you have more than 24 posts to your name Otherwise you will definitely develop a reputation as an annoying member, and so far you are going about that pretty well.

[mikeyakame]

All I can say is DDR2-1333 on a P45 is simply amazing. Those symmetrical read/write/copy speeds are also amazing. Nice work Cryptik

Also mate off topic of that, roughly what Clock Skews on A1/A2/B1/B2 did you find you needed to use to get the D9s to push over 1200Mhz on the P5Q Deluxe. I played around briefly trying to get my mates D9s on his Dominator 1gb sticks to push above 1150 on his P5Q deluxe and at least on the bios we were using then (1101 or something along those lines it was) had all kinds of issues where it wouldnt post what so ever at that frequency. Didn't have nearly enough time to properly figure out the necessary skews to get it to post in the short time I had with the board.