Turbo throttling is still present. So much for illuminatiASUS' statement on a new fixed bios for us.
Printable View
Turbo throttling is still present. So much for illuminatiASUS' statement on a new fixed bios for us.
There will always be turbo throttling on the generally released bios versions that you download from the Asus website. That's part of the original spec and they don't plan to change it. The special 0006 bios does not have any turbo throttling issues so if you need that feature then that is your only choice.
Would anyone know if they are going to make a bios so the memory multiplier can be unlocked like the RIIE ?
I asked illuminatiASUS that question, and he didn't answer. So my guess would be no.
Fair enough. Might have to go buy a Rampage II Extreme
@all
Is anyone know how to undervolted CPU PLL (< 1.80v) on this mobo (P6T/D Deluxe /v2).
i read it many times before, that undervolting "CPU PLL" might help in overclocking (lowering cpu temp thus gaining stability on 24/7 use oc. NOT in xtreme oc cmiiw).
& i eager to try it my self.
please enlighten me ( or send me a hacked turboV :D:D or special "undervolt" bios :D )
thx b4.
So what exactly does 0610 for the V2 resolve? I don't know if its worth the time to flash from 0504 (which has not given me a single issue thus far).
No real clue, aside from the G15 keyboard issue.
I found however that I can lower my IOH voltage to 1.20V now instead of 1.26V on my C0 920. Previously I would have gotten BSODs after 2-3 passes of Linx below 1.24V...
Anyone else experienced that phenomenon?
Edit: 0610 fixes the missing floppy controller support of 0606 too ...
I read that too. I guess it would take a hardware volt mod to shift the voltage down by a few notches compared to what you set in the BIOS. I'd do it, actually, if someone figured out which capacitor or resistor would ahve to be changed.
What did you read how far down they were going?
I get pretty good vcore for 4Ghz with the 0610 bios. ICH and IOH at 1.1v.
http://img15.imageshack.us/img15/2662/vcorek.jpg
well....usually i see 1.6v-1.7v (on various forum found by "google") but there is someone in this X's forum (of course, they named it "Xtreme" not for nothing :up:) can go down as far as 1.3v:shocked: (the lowest cpu *i7* PLL voltage known by me :shocked:)
http://www.xtremesystems.org/forums/...postcount=1181
is there any other way to do it ? ..beside "HARD MODDING" our motherboard.
Doesn't the Asus P6T Deluxe allow you to take CPU PLL down lower than 1.800v?
The UD5 starts at 1.300v and goes right up to 2.520v I think it is.
It was on EVGA forums that I read lowering CPU PLL can actually help to make your o/c stable, as I was having difficulty getting my 4.2GHz profile stable, so I tried using 1.300v CPU PLL, and it sorted everything out.
::edit::
Here is a screenie of my stable 4GHz o/c:
http://img171.imageshack.us/img171/8528/et6.jpg
ASUS doesn't listen E-CRUSH, that's why I just bought a clasified.
:D :D :D
L o L...
but,.. hey!! EVGA too ya' know...they wont sell their mobo's in my country when i told they so :rofl::ROTF: <-kidding, of course i never tell anything to evga.
so what can i do?? online petition ? :rofl: neh!...it will lead to more foe than friend. Beside that, Who will guaranteed it will make my oc beter :D
btw i wonder, it cannot be me ALONE (hope so) who like to have option for lowering CPU PLL..right ? ...or.....:shrug:
Out of curiosity to e_crush and the rest, do the Gigabyte X58 boards have Back to back CAS delay in their BIOS? I don't think it would be a good tradeoff if I had the lower CPU PLLs instead of B2B CAS...
AFAIK the ASUS boards were the only ones to support setting B2B CAS...
Else I'd be eager to try Gigabyte. The dirt cheap GA board in my second PC is a really good one for it's low price.
@amurtigress
I dunno, i never had GA's mobo before. Maybe you should ask mr. Bluntsmoker for this..
But we human, tend to be greedy,don't we? So it would be nice to have those two option in our beloved p6t series....do you agree?
Oh,btw..since you mention it. Will you tell me what is the optimal number of B2B CAS for 8-8-8-21@1600mhz..is there any formula for this option? Or maybe "auto" is good enough?
===================================
@all
Anyone care to tell me, the absent of <1.8v (on cpu pll) is caused by hardware limitation or it's only matter of time for us to have it in the future.
I love the P6T Deluxe :up:
Despite few shortcomings itsa nice mobo produced a max 230 BCLK with my 920 D0 just to boot in to windows, max stable BCLK seems like 225 BCLK though I guess it very from mobo to mobos...
Still looking for a way to reduce the PLL voltage from 1.80 as theres no option in bios :mad:
And any info on the proper clock skews for the 24/7 setting would be much appreciated. ;)
My 24/7 settings on air so far is...
http://img268.imageshack.us/img268/2582/4200mhz123v.png
me too .... i'am interested in clock skew (CPU & IOH) but i cannot find any meaningful article, only trial and error by someone else...
as i am is satisfied (and still try to learn it better :cool:) P6TD DELUXE user. But..if we can make it better why not?:D
ps : the GA mobo's can have B2B CAS delay via cpu tweaker program. (at least on their p55 series, dunno about their x58's mobos series)
this tread help me to understand it (B2B CAS delay) better :
http://www.xtremesystems.org/forums/...d.php?t=233960
(before mr "amurtigress" mention it yestreday, i had no clue, whatsoever on this powerful option)
and this :
http://www.madshrimps.be/?action=getarticle&articID=940
Hi E-crush
I found out that B2B CAS delay is commonly used by the few who know about it to increase stability on their RAM. Rule of Thumb: The higher the value, the more stability you usually get. My Corsair XMS3 modules have it programmed to 0.
Some testing showed me that along with the increased stability, if notable at all, you will lose bandwidth.
0-4: No decrease, hardly measurable.
6-8: Minor decrease
12: Memory bandwidth decreases to what you'd expect of a dual channel system. About 6-8 GB/sec
I am using a value of 4 these days.
I read some newer ASUS S1366 board has the value set to 12 fixed. I think it was in an article on the Rampage II Gene or so...I seriously wonder if that board loses bandwidth just as dramatically.
CPU PLL voltage:
There are two possible scenarios: The voltage controller chip for the PLL uses Voltage ID pins controlled by the BIOS. However, I cannot tell if 1.8V is the lowest that it can be programmed to, or if only the BIOS alone is preventing us from setting lower levels.
Second possibility, the analog part "Behind" the chip can be manipulated by adding a resistor to decrease the voltage by 0.2V or so, so 1.8V would refer to 1.6V on the board.
Either way, we are left guessing.
Hi there too! mr amurtigress :up: :D
First of all, i owe you a gratitude (a lot of...) for bringing in "B2B CAS" subject. For me it's the first time someone mention it..:D
I have been testing it too (playing w/ b2b) what i found is identical w/ yours finding and madshrimps article ( http://www.madshrimps.be/?action=getarticle&articID=940 )
but i wonder can you set yours b2b to < 4?...
mine (P6TD), "4" is the lowest number and "0" is the greyed number if we set it to "auto"....,at least... "0" & "4" i think is the same (bencmark result) as you mention it before.
and Dreaming it too...
they (asus) leaving it preoccupying my mind for 2-3 days....:rofl:
================================================== =================
@all
about clock skew...
mine set to 100ps (CPU) dan 0ps/normal (IOH)..my logic are (CMIIW please) :
As we overclock, the cpu (signal) will run/arrive faster than the rest of the system,,,so i delay it to making it "harmonious" :D :D with other signal (memory, pcie, ect...maybe..:shrug:)...my bencmark (wprime, vantage&06 cpu test, super pi...you name it) result is showing most stable (not faster) in this setting....making it >200/0 will make result (slightly) worse...or 100/100...200/100...200/200...will make result tend to inconsistent.
i never test it for oc stability ... because i my daily oc is only 3.927@1.224v (see what HSF i use to figure my "WALL" :ROTF: )
ps : if you already found my "wall"...(beside my knowledge and skill of course :D)....now you know why i most interested in "undervolting" LOL :ROTF: to save money you know....:rofl:
...at 4.28@1.288v it will BSOD in linx if the temp go to beyond 93c (i think..saw it many times just before BSOD)...but never hang in daily use / gaming though...still i don't like it....and trying to overcome it without buying new HSF...(but the upcoming NH-D14 is tempting me.....http://www.noctua.at/images/computex...d14.jpg....and or the prolimatech megashadow)
3x2Go gskill trident pc12800 C6 @ 950mhz 7 8 7 20 1T 1.65v
watercooling HK rev3 LT
http://img21.imageshack.us/img21/2580/4580.png
E-Crush,
as for making signals harmonious as you say (snychronizing) this is basically a GOOD idea. For example compensating longer signal runtimes to different RAM banks...
However I think in THIS case it is about avoiding signal interference by shifting them slightly.
A huge problem is for the 20x multiplier of our beloved 920s that so many signals derived from BClock are in exact sync. So rising and falling signals broadcast an electromagnetic pulse into neighboring traces on the board that run signals in sync. Flank meets flank and spike amplifies spike.
No, I gues sthe clock signal delays/skews are here to AVOID perfect synchronisation and thus spike adding/amplifying....
This whole thing is even more important on our previous S775 systems with it's rather analog AGTL+ bus
@amurtigress
Thanks for your explaination.
It will be a great addition for my knowledge
Hi thanx mate, sorry for the late reply, thanks for the infos and please do let me know if you can find any more useful articles, Im using 100 skew for CPU and normal for PCIE now but dont see much difference though. Therefore still in doubt what shoud settings when raising BCLK.