cancel
Showing results for 
Search instead for 
Did you mean: 

GTX 970/980 BIOS update for DisplayPort issues

yeasah
Level 7
There seems to be a lot of misunderstanding surrounding the updated reference Maxwell BIOS -- there are several threads inquiring on this topic but many of them are confused as to what the necessity of the BIOS update or whether it applies to the ASUS cards at all.

Background: many Maxwell cards have had significant issues with output on the displayport connector, exhibiting a black screen in various situations: on initial boot, after restoring from sleep, etc. Different driver revisions appear to express the problem differently, and more recent releases solve the problem entirely for some people -- but not all. The thread on the nvidia forums describing the issue is here https://forums.geforce.com/default/topic/777412/geforce-drivers/black-screen-on-windows-login-344-16... -- this link goes directly to the nvidia response, which indicates a two part solution to the problem: a new driver (344.80) and a new reference BIOS (84.04.31.00.00). The suggestion is that the driver *may* fix the problem alone, and if it does not that the new BIOS will be required.

In my case, I have an ASUS STRIX 970 connected to an ASUS ROG SWIFT -- so displayport is the only possible option for me. The problem occurs for me even with the latest driver or that specific hotfix driver, and the other suggested PCIe setting related hacks are not effective or applicable (I have a somewhat older motherboard, an ASUS P6T). If my system goes to sleep for long enough for the SWIFT to enter its sleep state (i.e. the ring light turns off), I can expect that when I bring the system out of sleep that the monitor will be indicating "no signal" and no amount of power cycling of the monitor, unplugging/replugging of the displayport cable, etc. will bring it back. A system reset is the only reliable way. Not to belabor the point: this is *really* inconvenient.

So, back to the BIOS. Running the ASUS gpu tweak software informs me that I have the latest BIOS for my card, which is:

Sep 19,2014 84.04.1F.00.2B

There appears to be some confusion (i.e. in this thread https://rog.asus.com/forum/showthread.php?54246-Asus-gtx-980-strix-new-vbios) as to whether this BIOS is actually newer than the one specified in the nvidia thread (84.04.31.00.00) -- the fact that the version number includes hexidecimal digits is perhaps part of the problem, maybe along with the fact that 1F in hex happens to be the same value as 31 in decimal. I will present two pieces of evidence that the currently shipping ASUS BIOS version is older than the version specified by nvidia.

First of all, the date. The nvidia post that advertises the newly available BIOS was in November 2014, but the ASUS BIOS was compiled much earlier than that, in September 2014. Second, let's look at the BIOS release history of a vendor with a more active policy of BIOS releases -- MSI in this case:

2014-09-10 84.04.1F.00.F1
2014-10-09 84.04.28.00.F1
2014-10-29 84.04.2F.00.F1
2014-11-19 84.04.31.00.F1

(This data is from the techpowerup database of BIOS: http://www.techpowerup.com/vgabios/index.php?architecture=&manufacturer=&model=GTX+970&interface=&me...)

Clearly there have been many releases of BIOS made available since September's 84.04.1F -- revision 84.04.28 in early October, 84.04.2F in late October, and finally 84.04.31 on the 19th of November: this is the revision that nvidia is referring to in their post about the resolution to the displayport problems.

I'm not demanding that ASUS match the frequency of firmware release of other manufacturers -- it's a point of competitive disadvantage, but I'm sure there's a cost to producing new releases and it's up to ASUS to determine what their policy is for this. But in this case, there is a firmware release that corrects a defect in the product that is causing real issues -- this forum already has several threads asking about BIOS releases and/or complaining about displayport issues: it is causing actual customers significant problems right now, including me.

Please, can we get a direct, non-evasive answer as to when we can expect an updated BIOS produced in November 2014 or later based on the 84.04.31 reference code for Maxwell-based ASUS graphics cards?
269 Views
187 REPLIES 187

Simbiat wrote:
You just expect everyone here to have a PG279Q with its unique function of "Deep Sleep"?..


I really dont know why you are dissing me.
I am trying to help people in my rare spare time, so if you dont have to say anything constructive, better not post.

@ Pokuz ... I will give your workaround a try ! I only saw your posts last night on the forum, but yea at this point in time I am willing to give anything a go.
Will give some feedback soon on whether or not it works. The deep sleep function is something I have disabled a while ago, but as for your monitor switching on timing thing, that's something I haven't tried yet.

Thanks for your response and trying to be helpful. 🙂

To the other horse sphincter's who claim "oh you know its an Nvidia problem" ... Wrong... you (yea you, who bought an ASUS product) are now at the mercy of Asus... Asus has an obligation towards their customers, to support the products they sell. If the technology is dodgy, and you are willing to take peoples cash in selling them, that's what you signed up to do...
Asus is paying Nvidia royalty licensing fees for Gsync support... if any feature doesn't work, Asus should be on Nvidia's back about this... I find it very hard to believe that a problem of this magnitude managed to get through any form of quality control. How did they not pick this up ???
Sadly if you want Gsync support, you have to use the display port at this point in time...
The lack of feedback or response from Asus on this matter is very dis-concerning... No feedback like "We are aware of the problem and we are working with Nvidia on the issue" ... just nothing.
Some of the other vendors are at least doing effort in trying to resolve this, hence my frustration with Asus technical support...
I feel, as a consumer that purchased a pretty expensive piece of kit, that we are left out to hang dry.
If its a hardware issue that cannot be fixed via firmware updates or software drivers, at least notify your customers what the plan of action is... or are you just going to leave it as is, and wait this blows over when the next generation vid cards / monitors are released, and hopefully the problem will be gone and forgotten ?

Pornolio wrote:
To the other horse sphincter's who claim "oh you know its an Nvidia problem" ... Wrong... you (yea you, who bought an ASUS product) are now at the mercy of Asus... Asus has an obligation towards their customers, to support the products they sell. If the technology is dodgy, and you are willing to take peoples cash in selling them, that's what you signed up to do...

The problem is with nVidia's chip, so it is nVidia's problem and all ASUS can do is wait for nVidia to find a way to fix this. And what magnitude you are talking about? A dozen or so users using DP? From factory perspective that's nothing. You have no idea how "creating a product" works, that no matter how you test something you still may miss something? No matter how rich a company is it has limited resources and it can't test its hardware in all possible configurations, only in a set of "average" ones. Why do you think ASUS wanted to get MY particular videocard? Because issue could be in that particular piece, that's why. Not in the one nVidia or ASUS tested prior the release, but in mine. And GTX980 did work fine compared to GTX970, the one sent to ASUS. Until a Win10 November updated came in. And that update did not touch the drivers.
So, please, do not spout nonsense like a no-brain-consumer, who blames shops for faulty hardware instead of blaming the manufacturer. Blame nVidia as the enthusiasts who thinks from manufacturing perspective. Or at least just as an enthusiast, who tries to see multiple perspectives of one problem. Especially, since the guys do seem to have found a solution they are testing now with nVidia.

Reading your replies just proves to me how blinded and uninformed you truly are.
Do yourself a favour, stop posting dribble, and go play in your room with your cuddly toys.

A few dozen users ? If you had half a clue of just how many users are having this problem... Its on support forums
from various hardware sites, company support forums, EVGA, MSI, ASUS, Nvidia, Operating system forums (Microsoft + Linux), anantech, tomshardware, Dell, the list goes on and on as new users get added to the list on a daily basis.

With regards to my comment on the quality control, which you again completely missed the point...
Read the following carefully and slowly... We are not talking about some small crappy anti-aliasing setting while playing some unheard of game written by a privateer in a basement. We are talking about something as stupid as switching a damn computer on and booting up.

Also as much as I can symphatize with vendors that its near impossible to test all possible hardware combinations, we are STILL talking about current Video card technologies with current generation screens. Again, the scale of the problem is not limited to one brand, its MULTIPLE different hardware vendors where this thing is sticking its head out. How does something like this go un noticed prior to launching ?

Look if I spent under a $100 for this kit, I wouldn't have bothered coming to the forums asking for help on it...
However this is high-end enthusiast level kit not doing something as basic as powering up when you press a button.
Yes, yes, there is a work-around where you can continue ripping out the power cables on a daily basis, till you get to the point where your connector inserts start wearing out, then you sit with a bigger problem which no software / firmware update will fix.

Do you get it now ? If you market something dodgy and you cant support it, the only thing that will surface from this... is brand damage.
Go read up on it... be a big boy.

I'm getting the exact same issues as Pokuz and Pornolio

Asus Strix GTX 970 + Asus MG279Q (FreeSync version)

It's like the displayport signal isn't "active" so when either device doesn't see a signal it just refuses to handshake. It's getting incredibly annoying, and like others, spent money on new cables and mobo thinking it was something wrong with my PC. Is this impossible to fix?

What I have to do right now is when I boot my PC just have to guess the small window in which they will read each other, then its perfect from there on. Is there no way we can tell the GPU to actively search for a displayport signal?

If you can't wake your monitor from standby here is the solution:
Disable the USB selective suspending option in Control Panel -> Advanced Power Options.

cheek
Level 9
had my sli 980 strix for over a year. still not fixed. asus came up with one vbios that didnt fix anything. i rma the monitor once. change cable once...i guess im resorting to switching off and on the monitor main power switch .
i believe this is nvidia poor implemnetation of the displayport 1.2. FYI i have the old swift. is it fixed in the new swift?

Simbiat
Level 7
Somehow I feel like signal loss mid-work started to occur more frequently on latest nVidia drivers...

Simbiat
Level 7
I feel at home as if I am at work, because we have similar issues with Intel's cards at work... This is so sad

x7007
Level 7
Does anyone has issue with Gpu Idle clock time that locked to 638 mhz randomly ?

970 after Idle stuck on 680 mhz clock max

http://forums.guru3d.com/showthread.php?p=5262578#post5262578

I opened thread and there are opened threads in Geforce forums.

https://forums.geforce.com/default/topic/916069/geforce-drivers/gtx970-clock-rate-throttling-and-tem...
https://forums.geforce.com/default/topic/916738/geforce-drivers/gtx-970-clock-speed-issue-with-361-x...

Can someone help with it ?