AFAIK if the system isn't coming out of sleep mode properly it's where your CPU isn't fully stable.
AFAIK if the system isn't coming out of sleep mode properly it's where your CPU isn't fully stable.
Not true. As mentioned earlier, several of us have had issues resuming from sleep on a high overclock. I still can't resume from sleep on a multiplier higher than 47x, but my OC is 100% stable at 49x (Linx, prime, gaming, encode, etc all stable). This issue still hasn't been fixed as of UEFI 1.61.Quote:
Originally Posted by Ket
People have mentioned in other threads that if you're using a multiplier that requires PLL Overvoltage, then standby will no longer function (a universal issue on P67 boards). Some CPUs can go higher than others without PLL Overvolt so it's definitely possible for some people to reach 5GHz with standby still working.
One of many threads that discuss this issue: http://www.xtremesystems.org/forums/...d.php?t=268285
The issue may not even be anything to do with manufacturers UEFI's but rather a quirk in the CPU architecture which would explain why the issue pops up at random multipliers. I'm steadily building a new bugs list but so far there really isn't anything thats noteworthy. All I have so far is;
- Can't always resume from sleep with a high OC (could be any number of things not board or UEFI specific)
- Some UEFI settings do not "stick" in saved profiles
Thats it :shrug:
I just stopped prime after 5 hours @ 5.0, and no errors. Am currently running multiple threads of memtest which is at 1200% coverage, and 0 errors. I would assume that the oc. is stable. Still wakes to black screen. I have experienced the black screen with unstable o.c's before, but with this system, even the mouse lights up where on the other systems it wouldn't.
I'm booting into 5 ghz without internal pll overvoltage, so that might be the difference...
How about Offset ~0.110-130 not working Ket? Also I've had the issue of fixed vcore not reading at what I've set (1.15v set, 1.35v sensor reading).
Can't comment on offset as I don't use it. But as for fixed voltage if I use LLC level 2 then set a voltage I get pretty much exactly that voltage. So if I set 1.36v with LLC lvl2 it'll be 1.36v idle in Windows and 1.352v with all 4 cores 100% load.
Its been a while in coming but I finally had time to finish it; Asrock P67 Extreme 4 review
Ket, have you tried running the board without the PCH heatsink cooler? Judging by your comments, it's more bling than cooling capacity. Since that's the case, any problems running the board without the PCH cover? Assuming a case with reasonable air flow.
As crap as the PCH heatsink is, its still required as the PCH does get reasonably warm. My advice would be to use a 3rd party heatsink but its been some years since manufacturers like Enzotech or Zalman have released heatsinks that would fit modern boards. I guess we are stuck with it.
I was hoping that the aftermarket south bridge cooler for the DFI Expert board that I have still kicking around would fit. I forget who the heck made it, but it was a vga cooler that fit what ever vid cards from that time period, Ti 4200's?
I think it was/is EverCool
http://www.svc.com/vc-re.html
lol.....that's the one! the good ole evercool! If you had an Expert, that was the first thing you did to it. Not sure how the pin holes of the Ex4 will line up on it though.
What is the point in switching from well working passive cooling to an active one?
The PCH heatsink on the Extreme4 ISN'T a well working heatsink. It's utterly sh|t. Theres no need for a active SB cooler though just a decent passive one.
Why is that? I have tested temps with pyrometer and result shows that it works fine as well as those radiators on power phase. How it looks it is just a matter of taste.Quote:
Originally Posted by Ket
@Ket - If I may ask few questions:Quote:
Originally Posted by Ket
1. What is potential benefit? Any stability issues possible?
2. What is the mounting of this heatsink? Will there be a (warranty void) visible evidence when alternative headsink would be used?
3. What is the mounting distance - would any of these cheap solutions fit:
a)http://www.zalman.com/ENG/product/Pr...ad.asp?idx=132
b)http://www.zalman.com/ENG/product/Pr...ad.asp?idx=131
I have a spare HR-05 SLI, but that seems like overkill for PCH :)
@Radier - what was the actual temp? You measured the chip or heatsink temp?
Just one more question - where is the MB temp sensor located on our boards?
There are basic design bases to adhere to when making a good heatsink, the PCH heatsink on the Extreme4 meets none of them. A basic and effective heatsink is simply a block with a good base and fins to allow proper dissipation of heat through natural airflow. The Extreme4s PCH heatsink has none of that and acts more like a heat trap than a heatsink, thus adding nothing but another "hotspot".Quote:
Originally Posted by radier
Basic Thermodynamics.
1. As you are stuck with the heatsink on the Extreme4 my advice is to lap the base to a nice finish, it'll help make it a bit more effective and of course use good TIM like MX3 or AS Ceramique.Quote:
Originally Posted by Yorick_Rise
2. Mounting of the heatsink is "standard", which is to say in my mind pretty poor. Its mounted with nothing more than tiny rubber feet on the heatsink to prevent it from wobbling and pushpins. I've replaced the pushpins with a couple small screws, washers and nuts for a much more secure fitting and to apply more pressure which will help make the heatsink more effective (more pressure = less air between heatsink and PCH)
3. I didn't measure the distance between the holes but I can say a heatsink with mounting holes spaced 54mm apart is too small.
As for warranty, just keep the stock mounting method handy should you ever need to RMA. Just replacing TIM and using a good mounting method won't kill a mobo.
Do you have any stability issues related to temps on PCH? As far as I remeber yours last actions with changing thermal compound on power phase the result was cold radiators which is not good effect.
Not cold, warm. Cooler than when the board used the crummy stock TIM and thats to be expected when moving from a mid-grade TIM at best to high grade TIM. Everything is making proper contact with the MOSFET heatsinks also, so nothing but genuine improvements.
I've just discovered something about UEFI 1.61 for the Extreme4 you might want to check for yourselves. I was gathering some idle and load temperatures for comparisons to the new cooler I have arriving and to my surprise IBT failed with settings I know are absolutely stable. So after a moment of a puzzled look I went into the UEFI and tweaked a few settings and tried running IBT again - still failed. I tried tweaking a few more things and still IBT failed so I reverted back to UEFI 1.6A and IBT has now passed with no OC settings changed.
I would advise caution with UEFI 1.61 and urge you to test settings that you know are absolutely stable with IBT to see if you also fail.
Ket, what's the difference between the shipping BIOS 1.60 & 1.6A?
1.6
1. Modify boot logo (can anyone say wtf is the point in that?)
1.6A (BETA)
1. Fixes on-off Bug when Spread Spectrum is disabled.
- Maybe decrease BCLK OC-ability.
I don't know if official 1.6 fixes the on-off bug, probably not as the "A" to me would suggest a UEFI more recent than just 1.6.
Ket, are there holes around the PCH southbridge, like there were on socket 478 or 775 boards?Quote:
Originally Posted by Ket
If there is, then zipties, my friend...zipties :D
You can easily frankenstein an old BX chipset heatsink on it and slap on a 30mm fan :) I even have an ic7 max 3 heatsink sitting around...
Can someone explain what "Core Current Limit" means and does?
http://img.photobucket.com/albums/v187/bizket/uefi1.jpg
"Modify boot logo", that is needed for B3 revision. They have a boot logo with "B3" in it.Quote:
Originally Posted by Ket