Sign in with
Sign up | Sign in
Your question

Best bios settings for Asus P8Z68-V Pro

Last response: in Motherboards
Share
June 8, 2011 5:30:40 PM

Hi all, I have the Asus P8Z68-V Pro mobo, i7-2600k and Corsair 2x4gb 1600mhz memory (CL8 or CL9). Can anyone suggest the best bios settings for this?
a c 207 Ĉ ASUS
a c 716 V Motherboard
June 8, 2011 5:40:19 PM

Q - HDD/SSD?
Q - RAID?
Q - What exact RAM?
Q - OC CPU?
Q - HSF?
Q - GPU and are you OC the GPU?
Q - Purpose e.g. Gaming, Desktop?
m
0
l
June 8, 2011 6:58:00 PM

no raid, just one HDD for now. i'm using the WD raptor 74gb. I also have a WD blue caviar that I will try using instead of the raptor.

the ram is at home so I don't have the part number for you. but is the corsair 1600mhz cas 8 or 9

I didn't OC anything.

i'm using the heatsink and fan that came with the 2600k

i'm using the onboard GPU and not OCing it.

purpose is desktop, internet. but I do want to play Diablo3, Duke Nukem forever and Dungeon Seige 3 when they come out.
m
0
l
Related resources
a c 207 Ĉ ASUS
a c 716 V Motherboard
June 8, 2011 7:45:44 PM

1. If applicable update your 'UEFI' BIOS {assuming it's not 0501}
2. BIOS:
a. Load Optimized Defaults
b. {assuming XMP RAM} ; AI OC Tuner -> X.M.P.
c. {verify SATA AHCI} ; SATA Mode -> AHCI Mode
d. Save and Exit = Yes

Optional, you 'can' use the EPU Switch on the MOBO PCB and 'Enable' the power savings feature; see pg 2-17.

Use the Blue SATA Port top/closet to PCB for the primary. SATA2 is the better choice for ANY mechanical HDD.

Good Luck!
m
0
l
June 8, 2011 8:26:41 PM

Ok I'll try those, thanks. but my HDD is already using a sata2 port. does XMP mode for memory mean the memory will run at a voltage higher than 1.5?
m
0
l
a c 207 Ĉ ASUS
a c 716 V Motherboard
June 8, 2011 9:15:04 PM

XMP simply means that it will run at its' Rated {Tested} Frequency, CAS Timings, Voltage; Example if the RAM is CMZ8GX3M2A1600C9B 1600 MHz, 9-9-9-24-2N @ 1.5v then those values will be set in the BIOS. http://www.corsair.com/vengeance-8gb-dual-channel-ddr3-...

However, if the RAM is 1.65v then 1.65v will be set in the BIOS. It has been proven the 1.65v is 'Urban Myth' to be harmful to P67/Z68 SB CPUs.
m
0
l
a b V Motherboard
June 8, 2011 9:20:15 PM

jaquith said:
However, if the RAM is 1.65v then 1.65v will be set in the BIOS. It has been proven the 1.65v is 'Urban Myth' to be harmful to P67/Z68 SB CPUs.


Oh really? Can you post a link for that finding? I find that hard to believe that we have proven in the space of 6 months since release that the Sandy Bridge CPU isn't damaged by 1.65v.

Maybe the processor can survive 6 months use 24/7 at 1.65v. All that proves is that the processor is usable for 6 months? Don't know about you, but most of us are buying Sandy Bridge to last 3+ years -- a full 600% greater than the amount of time that one could have possibly tested a SB chip.

An Intel rep on another thread recently re-confirmed Intel's position on the DRAM voltage, that position being that voltage greater than 1.5v+5% is above design tolerance for the Sandy Bridge CPU controller.

Here is the quote from June 1, 2011:

Quote:
A couple weeks ago I sent this question up to engineers to answer about the voltage if users can use 1.65v without any problems. The response that I got back was that it is possible to use the 1.65v memory but it is really pushing more stress on the processor’s memory controller then we would like and that if our Tech support finds out about you using memory at this voltage it can void the warranty. Now there maybe some setting in the Bios on a board that you can change the memory voltage down to 1.5v but I would rather tell you right off the bat to avoid it since most memory makers are offering both at the same price. Leaps-from-Shadows does a good job of covering this issue.

Christian Wood
Intel Enthusiast Team
m
0
l
a b V Motherboard
June 8, 2011 9:24:55 PM

velocci said:
Hi all, I have the Asus P8Z68-V Pro mobo, i7-2600k and Corsair 2x4gb 1600mhz memory (CL8 or CL9). Can anyone suggest the best bios settings for this?


I'm pretty sure that you and I are running the same memory. I would recommend that you use 1333MHz at 1.5v. If you go 1600MHz at at less than 1.65v, it may not be stable unless you are lucky enough to have above-average memory.

If you go 1600MHz at 1.65v, you risk damaging your processor. The risk-reward is not worth it -- you will not see the difference between 1333MHz and 1600MHz except in synthetic benchmarks. It will not noticeably improve your gaming frame rates and will not make applications load noticeably faster.

As I say in my previous post up, Intel recommends caution and suggests that they know who uses greater than 1.65v. I do agree that it's possible that:

1) Intel's recommendation is too cautious
2) Intel has no idea what kind of voltage you put through it without thousands of dollars of forensics research

But why take the risk of having to deal with the hassle of an RMA to put the memory at a speed slightly higher than stock that you'll never notice unless you're a benchmark junky?

I don't see a reason.
m
0
l
a b V Motherboard
June 8, 2011 9:49:57 PM

jaquith said:
Read through this post -> http://www.tomshardware.com/forum/290612-30-manually-vo...

Really 1.65v RAM is perfectly safe for Sandy Bridge! :D 


This is not a study; this is not something that I would base my decisions on. Instead, it is you and, Geekapproved, crewton, Noworldorder arguing back and forth. A few quotes:

From you:

Quote:
Truth be said, I see a lot more 'problems' with 1.50V RAM + SB. The link essentially says what I've stated, I too read both sets of Intel specs. I bought into the 1.65 'bad' until logic and documentation proved otherwise.


From GeekApproved:

Quote:
1.5v ram is what you want for P67. You can use 1.65v ram however, but I wouldn't recommend that if your overclocking at all. Anything over 1.65v can damage the cpu.


Quote:
I've not seen even one, out of hundreds of P67/H67/H61 that has a problem with 1.5v ram if the timings and speed is manually set correctly. Nor have I seen a single problem with 1.65v ram if the speed and timings are manually set correctly. However, Intel still recommends 1.6v max. Only time will tell if these overclockers start having cpu failures.


from Noworldorder:
Quote:
1.65v RAM is allowable on Sandy Bridge motherboards


From above, it's basically just arguments with no evidence, except the dueling recollections from you (I've seen more problems with 1.5v than with 1.65v) and Geekapproved (I haven't seen more problems with 1.5v).

The only real assertion based on any kind of facts is this gem of induction made by you:

Quote:
If you go strictly on Intel spec voltage then an most CPU OC would be impossible. DRAM Voltage up to 1.575v is 100% within spec.

After putting my nose into Intel's CPU manuals the LGA 1156, LGA 1366 {i7-900, i7-800, & i5-700} and LGA 1155 {i7, i5 and i3} ALL had the SAME "DDR3 I/O Voltage of 1.5 V" and SAME DRAM Voltage up to 1.575v.


Okay, so maybe it was fine for LGA1156, LGA1366 -- how does that mean that it works for LGA1155? It doesn't at all, because the last time I checked, Sandy Bridge is a pretty different animal than those other chipsets.

Here's your statement to the OP, who you are suggesting risk his expensive hardware to achieve a very minor performance improvement:

Quote:
However, if the RAM is 1.65v then 1.65v will be set in the BIOS. It has been proven the 1.65v is 'Urban Myth' to be harmful to P67/Z68 SB CPUs.


There is no proof. The only way in which the thread that you linked me to proves that it is an urban myth is you saying the following in that other thread:

Quote:
Therefore, this is all becoming more and more 'Urban Myth' that 1.65v RAM is 'bad' for the Sandy Bridge.


This, of course, is based on your questionable inductive argument quoted above.

To the OP: Stay at 1.5v
m
0
l
a c 207 Ĉ ASUS
a c 716 V Motherboard
June 8, 2011 9:54:14 PM

fullofzen said:
I'm pretty sure that you and I are running the same memory. I would recommend that you use 1333MHz at 1.5v. If you go 1600MHz at at less than 1.65v, it may not be stable unless you are lucky enough to have above-average memory.


Run it at Rated {Tested} Frequency, CAS Timing, and Voltage. DO NOT use 1.50v IF the RAM is rated for 1.65v @ 1600 MHz; this will cause errors! If the RAM is 1.50v @ 1600 MHz, most newer kits are 1.50v, then use 1.50v. Most RAM at 'SPD' Speed is 1.50v, but speed is 1066 or 1333 MHz depending upon the CPU. Sandy Bridge 'SPD' a/k/a default is 1333 MHz @ 1.50v.
m
0
l
a b V Motherboard
June 8, 2011 9:59:37 PM

jaquith said:
Run it at Rated {Tested} Frequency, CAS Timing, and Voltage. DO NOT use 1.50v IF the RAM is rated for 1.65v @ 1600 MHz; this will cause errors! If the RAM is 1.50v @ 1600 MHz, most newer kits are 1.50v, then use 1.50v. Most RAM at 'SPD' Speed is 1.50v, but speed is 1066 or 1333 MHz depending upon the CPU. Sandy Bridge 'SPD' a/k/a default is 1333 MHz @ 1.50v.


No, it will not cause errors. The Corsair memory the OP has will run just fine with 1333MHz at 1.5v while it is rated at 1600MHz at 1.65v. I have run this memory through memtest86 pass after memtest86 pass and with Prime95 Blend and have had no problems whatsoever at 1.5v and 1333MHz on my own machine.

There is a trade between voltage and memory speed. The higher you go on memory speed, the higher your voltage will need to be. Inversely, if you drop voltage down to 1.5v, all you have to do is reduce memory speed to a safe level.
m
0
l
a c 207 Ĉ ASUS
a c 716 V Motherboard
June 8, 2011 10:05:05 PM

fullofzen said:
This is not a study; this is not something that I would base my decisions on. Instead, it is you and, Geekapproved, crewton, Noworldorder arguing back and forth. A few quotes:

These 'types' of nested quotes are a little wackadoodle; don't do them.

G.SKILL specifically has 1.6v & 1.65v Sandy Bridge RAM http://gskill.com/products.php?index=357 & http://gskill.com/products.php?index=375

This makes complete sense and it's not my opinion alone, I went a step further as I showed in the linked post and thoroughly read the Intel technical documents. I too 'originally' had the same 'Urban Myth' until I read the Intel docs.
m
0
l
a b V Motherboard
June 8, 2011 10:07:20 PM

jaquith said:
These 'types' of nested quotes are a little wackadoodle; don't do them.

G.SKILL specifically has 1.6v & 1.65v Sandy Bridge RAM http://gskill.com/products.php?index=357 & http://gskill.com/products.php?index=375

This makes complete sense and it's not my opinion alone, I went a step further as I showed in the linked post and thoroughly read the Intel technical documents. I too 'originally' had the same 'Urban Myth' until I read the Intel docs.


Last I knew, G.SKILL is not a division of Intel. What do they know? They're trying to sell memory now, not sustain the life of an Intel processor.

Perhaps I misread your posts, but it sounded like you were saying that because Intel docs said 1.5v was the limit for 1366 and 1156 that since that was too conservative, it's probably too conservative for Sandy Bridge. Am I reading that wrong?
m
0
l
a c 207 Ĉ ASUS
a c 716 V Motherboard
June 8, 2011 10:08:24 PM

fullofzen said:
No, it will not cause errors. The Corsair memory the OP has will run just fine with 1333MHz at 1.5v while it is rated at 1600MHz at 1.65v. I have run this memory through memtest86 pass after memtest86 pass and with Prime95 Blend and have had no problems whatsoever at 1.5v and 1333MHz on my own machine.

There is a trade between voltage and memory speed. The higher you go on memory speed, the higher your voltage will need to be. Inversely, if you drop voltage down to 1.5v, all you have to do is reduce memory speed to a safe level.

This is bad advice.

IF the RAM is running rated 1600 MHz and IF it is tested at 1.65v --> running 1600 MHz at 1.50v RISKS needless errors.
m
0
l
a c 207 Ĉ ASUS
a c 716 V Motherboard
June 8, 2011 10:12:02 PM

fullofzen said:
Last I knew, G.SKILL is not a division of Intel. What do they know?

They KNOW a hell of a lot more than either of us! :lol: 
m
0
l
a b V Motherboard
June 8, 2011 10:14:40 PM

jaquith said:
This is bad advice.

IF the RAM is running rated 1600 MHz and IF it is tested at 1.65v --> running 1600 MHz at 1.50v RISKS needless errors.


Better that you have some temporary system instability while you find the best speed you can get between 1.5v and 1.575 volts than to pump up the voltage to unsustainable levels. You don't even have to boot into the OS to determine stability -- the OP can use memtest86 to assess the stability of the memory and have the hard drive totally unplugged to avoid respinning and respinning it.

And you're misreading my posts -- I'm not advising the OP that he run 1600MHz at 1.5v. I'm suggesting that he run 1333MHz at 1.5v. That will be totally stable.
m
0
l
a b Ĉ ASUS
a c 75 V Motherboard
June 8, 2011 10:22:35 PM

Guys flat out if you use DDR3 with a voltage of 1.65 you are pushing the tolerance on the memory controller on the 2nd generation Intel® Core™ processors. This can damage the processor and void your warranty on the processor.

So can you do it? yes it is possible to run a 2nd generation Intel Core processor with 1.65v memory but if you are going to be buying the memory new do yourself a favor and DON"T do it. Why take a risk for no performance gain throw away the money you spent on the processor in the first place?

Christian Wood
Intel Enthusiast Team

m
0
l
a c 207 Ĉ ASUS
a c 716 V Motherboard
June 8, 2011 10:46:28 PM

I trust Intel. Then the same would be true for the i7 9xx line QPI/VTT/VCCIO specs are same as SB.

What about G.SKILL SB 1.6/1.65 XMP RAM?!

And while I've got your attention, what's the highest safe vCore for the SB CPU?
m
0
l
a b V Motherboard
June 9, 2011 3:50:48 AM

1.65v is fine. according to the manual of my p8p67 asus delux board (also can be viewed on asus website) there are plenty memory modules tested at 1.65v and also 1.5v. If 1.65v was a nono then asus will not have included those memory modules into their Qualified Vendor List.


m
0
l
a b Ĉ ASUS
a c 107 V Motherboard
June 9, 2011 9:16:58 AM

The QVL simply lists memory modules that work when the manufacturer tested them. They do not test them long-term to make sure they still work after a month or six months or a year.

All Intel CPUs since Nehalem should have memory modules at 1.5v-1.575v only. Anything higher than that risks damage to the integrated memory controller. Also, anything higher than 1.575v voids Intel's warranty for the CPU. The "wink and a nod" from Intel engineers that up to 1.65v is fine does no good when your RMA is rejected because of what Intel considers overclocking/over-volting.

velocci:
If you bought Corsair Vengeance RAM, then the XMP settings will be fine. They will be set to 1.5v and the correct speed and timings.

I hope you're not planning to run Diablo 3, Dungeon Siege 3 and Duke Nukem Forever on the iGPU. You're going to need a discrete graphics card to be playable at any decent resolution and detail level for all three of those games.
m
0
l
June 9, 2011 1:30:54 PM

ok guys, i'm the original poster of this thread. this thread has gone off topic. i don't plan on using 1.65v. I want to stick with the 1.5v.

Fullofzen, where in the bios do you change the the frequency from 1600mhz to 1333mhz? and do you think this will my my PC faster?
m
0
l
a b Ĉ ASUS
a c 107 V Motherboard
June 9, 2011 4:26:21 PM

Before you go and reduce the RAM speed when you may not have to ... which RAM do you have? If you bought Corsair Vengeance RAM (like most of us recommend), you don't have to do anything except enable XMP. Corsair Vengeance RAM is designed to run at 1.5v anyway and the XMP will set it to that.

If you end up having to reduce the speed of your RAM, it will definitely not make your PC faster.
m
0
l
a b V Motherboard
June 9, 2011 4:58:34 PM

velocci said:
ok guys, i'm the original poster of this thread. this thread has gone off topic. i don't plan on using 1.65v. I want to stick with the 1.5v.

Fullofzen, where in the bios do you change the the frequency from 1600mhz to 1333mhz? and do you think this will my my PC faster?


We didn't go off-topic at all: you specifically asked if you would have to go above 1.5v for the XMP profile.

Quote:
Ok I'll try those, thanks. but my HDD is already using a sata2 port. does XMP mode for memory mean the memory will run at a voltage higher than 1.5?


To answer your question, though, the setting for this is in the "advanced" bios section under the a.i. overclock tab. A little ways down, there will be a memory speed selector; check 1333 there. A little further down, there is an entry for "DRAM voltage." Plug in 1.5v there.

Does that do the job for you?
m
0
l
July 7, 2011 9:04:44 AM

velocci said:
ok guys, i'm the original poster of this thread. this thread has gone off topic. i don't plan on using 1.65v. I want to stick with the 1.5v.

Fullofzen, where in the bios do you change the the frequency from 1600mhz to 1333mhz? and do you think this will my my PC faster?



Hi.

Saw your specs and found out we're have same components. And i see that we're both having the same problem. I just wonder i you have manage to properly tune your pc with the right bios settings maybe you can share those to me. I'm not sure if the optimized setting is enough for me.

Another things is that I tried rendering a simple scene with test render settings in 3dsmax and suddenly I'm warned about the cpu temp going 95 deg. Are you having the same problem with temp issues.

Thanks
m
0
l
September 18, 2011 5:15:29 AM

Quote:
Another things is that I tried rendering a simple scene with test render settings in 3dsmax and suddenly I'm warned about the cpu temp going 95 deg. Are you having the same problem with temp issues.


I have pretty much the same rig and I am having this exact problem. I have a P8Z68 - V Pro, 8gigs GSKILL 1600, i7 2600K. I went into BIOS EZ settings and set my machine to "ASUS Optimal" and every time I do a simple 3ds max render I get a temp warning. If I set my system back to the normal setting then CPU temp is fine, it doesn't go above 60 degrees C even for heavy renders. Can someone explain this behavior?
m
0
l
September 30, 2011 7:10:19 PM

This sounds to me like the 3D Max render engine is not tasking properly all cores in a new(er) multi-threaded environment. Some types of render engines keep a single thread task across BOTH cores in a earlier dual core system alignment (as in the last time they reworked the threading model of the software, and a bit of a poor coding man's hack i'll say). In a Quad core arrangement, this really stresses/utilizes only 2 of the 4 cores. This is a well known issue with older builds of 3D rendering software. Instead of being pipelined for ALL individual cores, it mearly passes along a SINGLE thread on up to TWO cores each if 2 or more are detected/reported.

To get around this, I've had to resort to using multi-pass rendering as way to get overheating under control. I use Lightwave 3D 9.5 now (it's now properly multi-threaded) but my older LW3D v.8.0 had this issue, but only on certain renders of volumetric lighting w/ or w/o animated/displaced textures that were not "baked" in the lighting passes yet, (kept a 100% load on only 2 cores, while the other 2 cores were idle). I found if I did the lighting passes separately, the issue went away (I don't know about 3D'S Max threading though). Kind of a bum deal to make separate passes for pre-vis, but really I prefer it into final renders where I need more lighting control for exact photoreal scene matching.

I suggest you expieriment with your various render settings and find the tool most responsible for the 100% load, then do that function as a separate pass if you can, to keep the loading in check. At some point an update will then "render" the issue moot and no longer be plagued by the "hack-style" CPU loading.

Hope it helps you out. :hello: 

m
0
l
December 12, 2011 2:32:00 PM

Huh??? :ouch: 
m
0
l
!