Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[LordBaal]

How much you paid for the graphic if you can share it.

[Mephisto]

Dunno. The whole thing was under 3k US dollarydoos after discounts, which comes in a few hundred under PC Part Picker's estimate. A bit south of 1600 or something.

[wierd]

Too rich for me...

I could only afford the ryzen 9, moboard, and a miniscule early installment on RAM.

GPUs just too pricy right now.

[LordBaal]

Is too crazy. Someone was offering 3080 cards here on 600 dollars. It seems too good to be true, probably is just a scam.

[Reudh]

Bought my first new PC since '13 and hoo boy this is going to be game changing.

For context, I'm coming from a 3rd gen i5, GTX 960, and 12 gigs of DDR3 something-or-other.



Not really computer advice, other than "go through a system integrator nowadays if you can."

phwoar, quite the rebuild! I did mine pretty recently (incrementally due to part prices) going from a 4th gen i5, 1060, and 16GB DDR3.
Currently using:

i5-10500 - it gets the job done, probably the weak point in the build right now
32GB Corsair Vengeance Pro DDR4RAM
Inno3D Geforce RTX 3080 TI (12GB VRAM) (this was like 65% of the rebuild cost right there, something like $2,400AUD), jumping from a clapped out mining 1080ti a mate gave me

[Lord Shonus]

Is too crazy. Someone was offering 3080 cards here on 600 dollars. It seems too good to be true, probably is just a scam.

Nvidia has a policy of brick in cards known to be stolen. So those might well be from a diveted shipment that turned out to be useless.

[Mephisto]

I'm pretty sure that's not a thing. As close they could get is marking them as stolen so the thief (or recipient of stolen goods) can't make warranty claims.

[Starver]

From a practical perspective, if there's a simple way for a hardware company to remotely-brick their products (intentionally!), then there's also a way for a huge Denial Of Hardware attack by everyone else and it'll be a Global Lockdown in a completely different sense. If it's a designed in capability (or internally discovered, but kept 'handy'), there's no way it'll not be leaked. And there are far more test-'til-it-breaks hobbyists out there, of no particular allegience, than any company's internal QR department could possibly coordinate for 'good'.

(Also, from a technical perspective, there almost certainly is* a way to accomplish the latter in such an apocalyptic way, it's just probably a rather tricky emergent bug that's not yet discovered (or used) by whoever might have actually worked out one exists.)

Not including the obvious method of planting bogus instructions on how to mis-reflash hardware (there are so many 'driver' sites out there that could be subverted) and getting the gullible end-users to break things themselves, just as careless ones already do.


* - still. There have been scares in this vein, in the past, patched up or otherwise mitigated.  And, no, not the one explaine ed in the Good Times warning.

[LordBaal]

Well is seems it ia a scam after all, we have MercadoLibre wich is the south american answer for amazon, and is very prone to scams. Also, I suspect they migth be second hand video cards used on custom mining rigs.

[Arx]

I've run into a bit of an odd problem with my GPU (GTX1060 3GB): it's not working as a full-on GPU, but it will run a screen (although the fan doesn't run, for some reason, which means it gets pretty hot pretty quickly, so I've taken it out while I'm not actively troubleshooting it). Checking the device manager turns up a code 43 (stopped because it reported an error), and checking the event log turns up that it failed migration (partial or ambiguous match).

The only thing I can think of that might have caused it is the most recent Windows update, but I haven't a clue how to resolve it. Updating the drivers doesn't seem to achieve anything -- in fact, I'm using GeForce Experience and it doesn't actually dismiss the "update available" after it installs which makes me question whether the update actually resolves. Reverting the Windows updates doesn't seem to have done anything either.

On the hardware side, the PCI power cable seems fine. Swapping it to the other PCI-E port doesn't help. I haven't tried it in another PC, but I may be able to later.

The precise series of events is this, repeating any time I update the driver:

Spoiler (click to show/hide)

Device PCI\VEN_10DE&DEV_1C02&SUBSYS_61603842&REV_A1\4&1479c31&0&0008 was not migrated due to partial or ambiguous match.

Last Device Instance Id: PCI\VEN_10DE&DEV_1245&SUBSYS_23101462&REV_A1\4&798c2c0&0&00DC
Class Guid: {4d36e968-e325-11ce-bfc1-08002be10318}
Location Path:
Migration Rank: 0xF000FFFFE000F122
Present: false
Status: 0xC0000719

Spoiler (click to show/hide)

Driver Management has concluded the process to add Service nvlddmkm for Device Instance ID PCI\VEN_10DE&DEV_1C02&SUBSYS_61603842&REV_A1\4&1479C31&0&0008 with the following status: 0.

Spoiler (click to show/hide)

Driver Management has concluded the process to add Service NVDisplay.ContainerLocalSystem for Device Instance ID PCI\VEN_10DE&DEV_1C02&SUBSYS_61603842&REV_A1\4&1479C31&0&0008 with the following status: 0.

Presumably the exact IDs are only useful to me, but there are the details.

Anyone have any ideas? I can't find anything more than what I've mentioned so far because of the glut of "ez fix" articles aiming for a high page rank to run ads.

[feelotraveller]

This page https://nvidia.custhelp.com/app/answers/detail/a_id/3075/~/windows-has-stopped-this-device-because-it-has-reported-problems(code-43)-for suggests some troubleshooting ideas.  If that doesn't work (before replacing the card) follow the link from that page to the microsoft windows error codes page which has an alternate troubleshooting suggestion.  Maybe one of those will help.  If not it might be worth posting the problem to one or both companies and see if they can offer anything more concrete.

[LordBaal]

Uh oh... should not the fans start regardless of whatever.. checked them to see they are not stuck?

[delphonso]

Uh oh... should not the fans start regardless of whatever.. checked them to see they are not stuck?

Some GPUs don't spin up unless they start to heat up past a certain threshold - I guess to reduce noise when not gaming (which would drown out the rest of the noise, presumably)

[Arx]

Uh oh... should not the fans start regardless of whatever.. checked them to see they are not stuck?

I think it's a software problem. The fans spin up when the PC powers on and spin back down when Windows boots, which is a little odd.


I can now add updating BIOS to the list of things I've tried that haven't apparently fixed it.

[wierd]

This is Windows being a fucking nanny.

Mittens mode on.

All that horse shit.


In english: 

Windows thinks the PCI ID for that card is ambiguous for more than one device in its driver cache, and cant figure out which one to install/enable.

Rather than simply ask you, Microsoft has taken to the belief that thier end users are beyond clueless, so to protect the hardware from the ambiguous ID (and loading potentially the wrong driver), as well as to protect the OS from an idiotic end user, it just decided for you that your GPU should use the SAFE!! generic VESA driver instead.  You dont actually NEED those 3D capabilities to display a desktop, so it all fine!

Now...

There  MIGHT be a way to force the driver to install, by uninstalling all prior installations of GPU software, including Experience, removing the device from the device manager, rebooting, then installing fresh.

If, after that it STILL does not work, you can try disabling driver enforcement, then MANUALLY updating the driver. (Windows will bitch quite a lot during that process though.)