• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

New ASUS BIOS Updates Released Today! Possible Skylake Bug Fix!

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

MNMadman

Member
Joined
Dec 13, 2015
Get 'em while they're hot! My M8G updated to 1402 without problems, though I didn't expect any.
 
Last edited:
OP
MNMadman

MNMadman

Member
Joined
Dec 13, 2015
...for those who might be waiting for a BIOS update to get rid of the Skylake bug, yes. Usually people are on top of this kind of thing.

Edit: Fixed the thread title. Does that make it clearer for you?
 

Woomack

Benching Team Leader
Joined
Jan 2, 2005
I'm just wondering if this BIOS is changing anything as it was announced 28 December while release of previous version. The same for all M8 boards.

On the other hand skylake issues are related to Prime95 and only small % of users is actually using Prime95. I haven't seen any issues even in Prime95 so it's not even something that always appears but may happen.
I only count that asus improve memory overclocking as now is about 1-2 ratios behind asrock or msi.
 

tachi1247

Member
Joined
Oct 28, 2008
...for those who might be waiting for a BIOS update to get rid of the Skylake bug, yes. Usually people are on top of this kind of thing.

Edit: Fixed the thread title. Does that make it clearer for you?

Yes, it does make it clearer. Your original thread title left out the most important part of the message you were trying to convey. It would be like creating a thread titled "Samsung made a press release today" and then in the first post only saying "I read it, it was good."

Anyway, back the the topic of the thread, the description for the new BIOS release doesn't mention anything about addressing the bug which leads me to believe it doesn't. It certainly would be worth noting since it can create large problems for commercial or institutional users (not that they would be using an M8G). Chances are this bios release was in the works far far before Intel announced the bug last week and was just routine improvements to stability as noted.
 

EarthDog

Gulper Nozzle Co-Owner
Joined
Dec 15, 2008
Location
Buckeyes!
Im still confused... LOL!

First, +1 to Woomack..

Second, was this for all boards in the Z170 line from ASUS or just the M8G?

EDIT: The Extreme got it... listed as: Update Microcode and improve system compatibility.

Whatever that means. I didn't run across the P95 bug either when I tried, so, I can't tell if this fixes it or not. Seeing as how it appears to be across all models, and it was stated that a microcode update would fix it, I would imagine it would fix that.
 
Last edited:

mimart7

Member
Joined
Jun 6, 2011
Im still confused... LOL!

First, +1 to Woomack..

Second, was this for all boards in the Z170 line from ASUS or just the M8G?

EDIT: The Extreme got it... listed as: Update Microcode and improve system compatibility.

Whatever that means. I didn't run across the P95 bug either when I tried, so, I can't tell if this fixes it or not. Seeing as how it appears to be across all models, and it was stated that a microcode update would fix it, I would imagine it would fix that.


There is a "bug" that affects Skylake when doing complex workloads, http://arstechnica.com/gadgets/2016/01/intel-skylake-bug-causes-pcs-to-freeze-during-complex-workloads/
 

EarthDog

Gulper Nozzle Co-Owner
Joined
Dec 15, 2008
Location
Buckeyes!
Yep.. I know. :)

As Woomack said, it tends to show itself with P95 and stress testing.. otherwise, its a non issue for most people... doesn't even effect all when running P95.


uses Fast Fourier Transforms to multiply extremely large numbers. A particular exponent size, 14,942,209, has been found to cause the system crashes.

I was confused about the thread/if there were BIOS updates for the entire ASUS line... that's it. :)
 

Woomack

Benching Team Leader
Joined
Jan 2, 2005
Even if error appears then it doesn't mean it will cause application crash or something. Stress test won't let you pass as it will find error but other applications will rerun operation or will pass to next task ( not all are crucial ). There is error correction in modern systems so if anything happens then it's often recalculated/corrected. Check system monitor in any Windows. There is info about actual calculation errors. Even fully stable system has some.
On the same way in the past there were 30+ potential errors while calculating on Prescott CPUs or 9+ on Northwoods. I don't remember other series but each generation has something. Actually Intel has never released any error-free CPU but not all data is public.

Each generation of CPUs/chipsets has some issues. 6 gen = chipset issues ( B2/B3 stepping replacements ), 7 gen = delays and problems to stabilize chipset what caused much weaker X motherboards than they should be, 8/9 gen = issues with power states, USB and memory support, 10 gen = so far only that weird cpu bug so can say it's pretty successful.
 

mimart7

Member
Joined
Jun 6, 2011
Errata in cpu's is nothing new. But a software developer buddy had plans to use Skylake in a production machine, but now they cannot.
 

mimart7

Member
Joined
Jun 6, 2011
So he has tested it and has errors with his work?

No he hasn't but because some of the work he does can put a heavy load on the cpu, he won't buy one. This a quote from him on another forum:

"We run into math errors that lock up our application more than we should. I think we have 3 documented complaints to Intel in the past 5 years."
 

EarthDog

Gulper Nozzle Co-Owner
Joined
Dec 15, 2008
Location
Buckeyes!
It's only the fft loads though...

So he had locks up in the past 5 years too with non skylake...he won't purchase it without knowing if it affects his work or not. Sounds logical. :p

It's not they cant...they don't know amd didn't test. It's that they WONT. :)
 

RJARRRPCGP

Member
Joined
May 30, 2004
Even if error appears then it doesn't mean it will cause application crash or something. Stress test won't let you pass as it will find error but other applications will rerun operation or will pass to next task ( not all are crucial ). There is error correction in modern systems so if anything happens then it's often recalculated/corrected. Check system monitor in any Windows. There is info about actual calculation errors.

I don't have errors like that and BTW, Windows-correctable errors can't be detected by an application, only the kernel.

And anyone who tells someone that hardware errors are normal, are probably noobs that want the big 5.0 or the like and is in denial that the CPU core batch can't handle the big 5.0 or the like...

If you're at stock or not a greedy OC'er, you probably won't get any hardware errors. Simple as that.

Hardware errors are 99 percent a component being pushed too far, cheap PSU, bad caps or bad solder joints.
 

Woomack

Benching Team Leader
Joined
Jan 2, 2005
Single errors are normal. I don't know what are you doing for life but I work in IT for many years and I've seen these errors on various platforms: servers, desktops, laptops ... no difference.
I also don't know if you remember Intel declaration about possible hardware errors in their CPUs. As I mentioned there was a lot of noise about it after Prescott release. Every hardware has that but it's usually corrected at different levels.

This is Q6600 @stock which is 100% stable and runs like that for ~5 years ( just before was Win7 ). Polish Win10 but you can see what is system showing.

1.jpg

This is stock i5 6300HQ in ASUS laptop which I got ~2 months ago:

4.jpg

If I wait some time after application start then errors drop to 0 but just after loading any application there are single errors.

If there is any uncorrectable error and is repeating then application or OS will crash. Usually critical errors end at blue screen.
 
Last edited:

Johan45

Benching Team Leader Super Moderator
Joined
Dec 19, 2012
Right if it ain't broke don't fix it. These types of CPU corrections always make me wonder about performance loss and AFIK once it's done it can't be undone can it?
 

Woomack

Benching Team Leader
Joined
Jan 2, 2005
If you have too many corrections then performance is lower because CPU has to calculate it again. You can see that if you overclock something too much and it's passing benchmark but score is worse.
1-200 errors is usually nothing comparing to couple of million operations.
 

wingman99

Member
Joined
Dec 10, 2003
It's hard to believe that a CPU will run with errors in some sections of the CPU because they are ones and zeros, Binary code with X86 architecture, 64 bits out-of-order execution , programs can round up and down and do error checking to go to the next step, cpu's can't. however I think Floating-point unit causes allot of trouble in rounding division up up up or down down. +1 for RJARRRPCGP, When Intel or AMD tests the transistors operating frequency for reliability they do it with a safety margin, so when you run the CPU at OEM settings also memory you won't have a problem with lockups.

When you do math calculations with a CPU and software like Pime95 or large math calculations there is no way to loop around in a software programing tree to finish the math without a error or crash, if there is a bug in the CPU.

The Pentium FDIV bug was a floating point bug, though rarely encountered by most users (Byte magazine estimated that 1 in 9 billion floating point divides with random parameters would produce inaccurate results), both the flaw and Intel's initial handling of the matter were heavily criticized. Intel ultimately recalled the defective processors.
https://www.coursehero.com/file/12316974/The-Pentium-FDIV-bug-research/

QUOTE:The trouble with Floating point numbers

We all know of floating point numbers, so much so that we reach for them each time we write code that does math. But do we ever stop to think what goes on inside that floating point unit and whether we can really trust it? I hate to cast aspersions on its good name but when I hear stories of space craft crashing, inconsistent … http://forums.theregister.co.uk/forum/1/2006/08/12/floating_point_approximation/#c_353

It's a good thing there is Prime95 program around, it is a good program to test for integer x86 and FPU also instructions. Drug company's, Science, Researcher, Engineering use integer FPU a lot.

What I can't believe this day and age some people are ok when we find flaws now called a bug, it was not like that in the Pentium days the public spoke out and there was a recall.
 
Last edited:

Johan45

Benching Team Leader Super Moderator
Joined
Dec 19, 2012
These eratta bugs have been around forever. Look at the original phenom they released a patch for it and for the most part it just killed performance to fix the problem which only happened sometimes.