Halp with 290x problem

Computer had stayed up all night although what I didn't know at the time was the slot is a PCI express at 4x. I returned the card to the PCI 3.0 16x slot and ran into issues again, this time starting as early as the boot selection screen for Win 10. This is starting to look like a motherboard issue of some kind. Later today I'm going to be trying out my old 270 in the PCI 3.0 16x slot.

Update: Reseated the CPU, no change. Used BIOS switch on the card, no change. without a spare card on hand, seems getting access to it today isn't going to happen, testing the slot itself is unpossible.

I'm debating if updating the bios on the motherboard is worth the time and effort. Also debating what to do moving forward, do I switch over to Ryzen when it releases or give up on AMD and go with an i5 mid range and a decent motherboard. I'm starting to question Asrock's quality based off some of the things I've been seeing over the last few days.
 
Last edited:
Why is updating the BIOS an issue?

Update the BIOS then reset your mobo by removing the battery and shorting the pins.

If that does not work, then the mobo is bad...
 
Mostly cause it takes time and effort beyond breathing. I've had a few weird things with this board from day 1 and as most of my best computering days were (I think) before so much stuff moved onto the CPU, I bring another point of failure to the table too.

In the beginning I had some RAM that wouldn't run at its rated 2133 speeds, anything beyond 1866 wouldn't boot. Eventually a forumite sent me some AMD RAM and it worked like a champ.

It comes down to a value proposition, I'm gonna grab a 570 GTX from my gf tonight to try in that slot and see if it'll do 2.0 x16. If it does work what does this mean, if not then back to bad CPU or board. Which then means either using hardware at less than optimal performance for some amount of time or again saying screw it and abandoning this platform for what would likely be an i5.
 
Ah OK... I must say I am still on a Sabertooth by Asus after having 5 straight Asus boars and never had an issue, not even once.

Back in 2004, I did get one that started acting weird and shorting the pins for a hard reset was all that was needed...

Fun fact: by 3 previous mobos are STILL running in various family member's PCs...
 
MasterGoa said:
Ah OK... I must say I am still on a Sabertooth by Asus after having 5 straight Asus boars and never had an issue, not even once.

Back in 2004, I did get one that started acting weird and shorting the pins for a hard reset was all that was needed...

Fun fact: by 3 previous mobos are STILL running in various family member's PCs...
Click to expand...


I tried an Asus motherboard when 133 MHz ram was fast. Nothing but headaches. This was also before their sabertooth line. I guess I'll try the cmos after the GTX 570. I didn't realize virtually any i5 just obliterates the a10 in everything. :(
 
Last edited:
That card does draw some power probably around 300w when fully loaded. Not sure if you tried but in the drivers under Games, turn the Global Powertune all the way down in the drivers, maybe even underclocking both the gpu clock and memory. I wondering if the PCIe power from the board is the issue here.

Board may work fine with a lower power card in that slot. Getting that card to another PC if available maybe can help clarify source of problem.

Looking at the manual, assuming a Fatalty FM2A88X+, I don't see where in the bios you can disable the CPU GPU, if you can I recommend doing that. You can do that in the Windows device manager to disable the APU GPU if it is turned on. Just to eliminate a possible conflict.

Also under North Bridge Configuration in the bios screen make sure PCI Express is selected for Primary Graphics Adaptor.

In the bios Storage Configuration - Turn on AMD AHCI Bios Rom to see if that makes a difference.
 
Last edited:
Replace the board. It's not the CPU causing PCIE failure. Sounds like the board has been bad for a while (maybe bent pins on the socket?) and now it's starting to eat ****.

RMA time
 
noko said:
That card does draw some power probably around 300w when fully loaded. Not sure if you tried but in the drivers under Games, turn the Global Powertune all the way down in the drivers, maybe even underclocking both the gpu clock and memory. I wondering if the PCIe power from the board is the issue here.

Board may work fine with a lower power card in that slot. Getting that card to another PC if available maybe can help clarify source of problem.

Looking at the manual, assuming a Fatalty FM2A88X+, I don't see where in the bios you can disable the CPU GPU, if you can I recommend doing that. You can do that in the Windows device manager to disable the APU GPU if it is turned on. Just to eliminate a possible conflict.

Also under North Bridge Configuration in the bios screen make sure PCI Express is selected for Primary Graphics Adaptor.

In the bios Storage Configuration - Turn on AMD AHCI Bios Rom to see if that makes a difference.
Click to expand...


All of those settings are correct, I recently disabled the onboard in the bios. I had it disabled in Windows for forever while I had the 270.

BIOS set for PCI-E has been a thing for a while.

ACHI is already on.


Nunz said:
Replace the board. It's not the CPU causing PCIE failure. Sounds like the board has been bad for a while (maybe bent pins on the socket?) and now it's starting to eat ****.

RMA time
Click to expand...

RMA isn't possible, the system is a full 3 years old and change at this point. If it's time for new gear it's new motherboard and CPU, trying to avoid new RAM.
 
So many mobos now come with 'reinforced' pcie slots. Mobo makers must know it was an issue.

Asus mobo I have upstairs is 7 years old and still kicking. I hope this gigabyte does as long a service. Stick to name brands theres no point in saving a couple bucks on a mobo.
 
Cleared CMOS, and that did nothing. Put the 270 back in and it ran like a champ as did a GTX 570. I have a feeling at this point that the card is at fault although my motherboard has added a degree of grumpiness too, the BIOS doesn't detect a second stick of RAM where it did before plus HWInfo64 does once Windows is booted. The main screen specifically, the onboard graphic system layout shows it being there, when I go into the ram timings both sticks are there too. Windows itself doesn't detect things, other applications do.

The ASrock board I went with received raving reviews hence choosing it. I have a feeling that this was a mistake.

PS Yes I did try both sticks in the 3 available slots, one of them is covered by my heatstink.
 
Last edited:
You said earlier it ran fine in a different PCI-E slot? Why not just run the card in that slot? Even if it's 4X, I don't think it should affect performance that badly. Also, if the card is stable in that slot it makes me feel like it's less likely that the card itself is bad.

You could also try reducing the speed of the PCI-E x16 port to x8 and see if that helps.

Here's a review that looks at the difference:

http://www.tomshardware.com/reviews/pcie-geforce-gtx-480-x16-x8-x4,2696.html

Basically, in some games, no effect. In other games there is an effect, but it's only about 10%. All things being equal running at x4 or x8 instead of x16 isn't optimal, but if you're in a situation where only x4 will work then there's still only a small penalty involved. If it really worked in the other slot, then I'd focus on at least getting it working, and then worry about changing things up later.

Edit: I also don't understand why you won't update the BIOS. It takes like 5 minutes to do. You should be able to even do it using a utility from Windows. Believe it or not, BIOS updates do fix things. Back in 1999 I had a problem where my motherboard would have all keyboard input freeze after being on for a couple minutes in a game (I remember the time period because it happened constantly when running EQ), and after updating the BIOS everything worked fine. Updating the BIOS is one of the first things you probably should have tried. It may not fix the problem, but there's a chance it could.
 
Last edited:
pax said:
So many mobos now come with 'reinforced' pcie slots. Mobo makers must know it was an issue.

Asus mobo I have upstairs is 7 years old and still kicking. I hope this gigabyte does as long a service. Stick to name brands theres no point in saving a couple bucks on a mobo.
Click to expand...

ASRock is a name brand. I've never used them myself (I've stuck to Gigabyte since EPoX and ABiT died), but they're considered legitimate. FWIW, you can read bad things about every manufacturer.
 
Last edited:
ASCI Blue said:
Cleared CMOS, and that did nothing. Put the 270 back in and it ran like a champ as did a GTX 570. I have a feeling at this point that the card is at fault although my motherboard has added a degree of grumpiness too, the BIOS doesn't detect a second stick of RAM where it did before plus HWInfo64 does once Windows is booted. The main screen specifically, the onboard graphic system layout shows it being there, when I go into the ram timings both sticks are there too. Windows itself doesn't detect things, other applications do.

The ASrock board I went with received raving reviews hence choosing it. I have a feeling that this was a mistake.

PS Yes I did try both sticks in the 3 available slots, one of them is covered by my heatstink.
Click to expand...

If the card ran good on the X16 but PCIe X4 slot then internally the card is good as in GPU/Ram. Sounds like the card is not getting sufficient power from the PCIe 3.0 X16 slot or damage on the card PCIe contacts but then that would show up in the other socket.

Maybe take that card down to a computer repair shop and put her in another machine to see if the problem goes with it. If the card is good then you could also sell it and get yourself a new motherboard. Not knowing if the card is good or bad I wouldn't sell it yet.

If the card is bad it is most likely still under warranty and let ASUS get you a new one or a replacement.
 
Updating the BIOS isn't going to do anything. You have a bad board, just replace t already. Return the PSU you bought if you need the cash, because that wasn't the problem.
 
The reason I suggested the BIOS reset and upgrade is because I was faced with a very similar case.

There is all but 5 minutes of work in updating a BIOS and resetting it...
 
Nagorak said:
You said earlier it ran fine in a different PCI-E slot? Why not just run the card in that slot? Even if it's 4X, I don't think it should affect performance that badly. Also, if the card is stable in that slot it makes me feel like it's less likely that the card itself is bad.

You could also try reducing the speed of the PCI-E x16 port to x8 and see if that helps.

Here's a review that looks at the difference:

http://www.tomshardware.com/reviews/pcie-geforce-gtx-480-x16-x8-x4,2696.html

Basically, in some games, no effect. In other games there is an effect, but it's only about 10%. All things being equal running at x4 or x8 instead of x16 isn't optimal, but if you're in a situation where only x4 will work then there's still only a small penalty involved. If it really worked in the other slot, then I'd focus on at least getting it working, and then worry about changing things up later.

Edit: I also don't understand why you won't update the BIOS. It takes like 5 minutes to do. You should be able to even do it using a utility from Windows. Believe it or not, BIOS updates do fix things. Back in 1999 I had a problem where my motherboard would have all keyboard input freeze after being on for a couple minutes in a game (I remember the time period because it happened constantly when running EQ), and after updating the BIOS everything worked fine. Updating the BIOS is one of the first things you probably should have tried. It may not fix the problem, but there's a chance it could.
Click to expand...


There is no way in the BIOS that I could find to run the slot at a slower speed.

Forgot to mention earlier that I did update the BIOS to the latest, part of the reason I couldn't is 1. I couldn't find any of my flash drives. 2. ASRock doesn't officially support Win 10 installs of the BIOS for the motherboard I have, it specifically says on the bios part of the website to use DOS or the upgrade via the BIOS.

That article is also 7 years old at this point, what was true back then may not be true now. I'd like to see a more recent piece with newer hardware.


At this point I think I'm with Nunz, either the board or CPU is bad. I repasted/reseated it with no success and for whatever reason parts of the BIOS see my 16 gigs of RAM, other parts done and the same goes for Winders 10. Dropping the CPU and bending a dozen pins didn't help either I'm sure. I'm looking at GPU-z now with the 270 in and it says it's a 3 x16 slot running at 2 x16 speeds. It also has the incorrect RAM slot being filled listed as 2 and 3 when it's really 2 and 4.

Yes I tested both sticks out in as many slots as I have access to, the CPU fan blocks one. Both sticks work fine, it's the lower two slots that are non-functional.

ASRock seems to be a legit brand, although I'm wondering if they're like Abit of yesteryear. Abit from my experience and several others I know made great boards..it often times took a few tries to get one of them.
 
If your girlfriend gave you a 570 GTX does that mean she has a computer that you can try the card out in? That might be a good step to try to isolate the problem.
 
Nagorak said:
If your girlfriend gave you a 570 GTX does that mean she has a computer that you can try the card out in? That might be a good step to try to isolate the problem.
Click to expand...


I tried it at her house today. Both slots on my motherboard are listed as PCI-e 3 x16, one only runs at PCI-e 2 x4 though. Results follow:

290x doesn't run stable in PCI 3 x16 slot
290x runs stable in PCI 2 x4 slot

270 runs stable in PCI 3 x16
270 untested in 2 x4 slot

GTX 570 runs stable in PCI 3 x16 slot (at 2 x16)
GTX 570 untested in 2 x4 slot.


In HER machine: i5 2500, MSI motherboard, 12 gigs of RAM, Corsair CX 750 can be seen here: https://www.msi.com/Motherboard/H67MSE43_B3.html#hero-overview

290x Doesn't run
270 Doesn't run
570 GTX runs like a champ

Both of the Radeons will allow the fans to spin up on the CPU, case, and GPU but the system will NOT post. I cleared the CMOS and tried with no success. I reinstalled the GTX 570 and had to clear the CMOS a second time to get it to POST and boot.
 
Wow, you're not having much luck. I've never had a graphics card flat out refuse to post in a board.
 
You must log in or register to reply here.
Back
Top