Monday, February 24th 2025

NVIDIA GeForce RTX 5080 Spotted with Missing ROPs, Too

We previously covered that NVIDIA's GeForce RTX 5090 and RTX 5070 Ti chips were spotted with a few missing ROPs. NVIDIA confirmed this issue affects 0.5% of the supply of GeForce RTX 5090 / 5090D and 5070 Ti GPUs, and users should contact their vendors for a replacement. However, the case of missing ROPs is now extended further, with one user on Reddit reporting that his latest GeForce RTX 5080 Founders Edition GPU is reading only 104 ROPs, instead of the regular 112. That is eight missing ROPs—a number similar to eight missing ROPs found on GeForce RTX 5090 and RTX 5070 Ti. To double-check, the user installed the latest version of NVIDIA drivers and still recorded 104 ROPs only.

We previously found eight missing ROPs constitute a 4.54% loss in the GPU's raster hardware capability. The GPU's Raster Operations Pipeline (ROP) units handle the final stages of graphics processing—they manage pixel-level operations like blending colors, smoothing edges through antialiasing, writing pixels to texture memory, and ultimately storing the final pixel values in the frame buffer. This differs from shading units (GPU cores), which focus on calculating the colors, lighting effects, and material properties of pixels and vertices during rendering but don't directly work with the frame buffer. The performance loss from missing ROPs will differ from game to game, depending on whether a game heavily relies on ROP-intensive operations. Nonetheless, with GeForce RTX 5080 FE in the picture, we must wait and see if more future SKUs may appear with missing ROPs.
Update Feb 25th:
In response to this discovery, NVIDIA provided the following statement to TechPowerUp
NVIDIAUpon further investigation, we've identified that an early production build of GeForce RTX 5080 GPUs were also affected by the same issue. Affected consumers can contact the board manufacturer for a replacement.
Sources: Reddit, via VideoCardz
Add your own comment

60 Comments on NVIDIA GeForce RTX 5080 Spotted with Missing ROPs, Too

#26
docnorth
SIGSEGV"You can easily ask for a replacement." -- I like this comment. :roll:
That's correct, you can easily ask for a replacement. When you are going to get it is a different story:nutkick:.
Posted on Reply
#27
piloponth
ImoutoYou know they are going to stop reporting them in the driver like they did with hotspot temps.
They can, but it will not save them. You can always run generic benchmark(s) and compare. If they block readings, some nice folks will write a specific benchmark/test to determine ROPs number by running performance test. The same as memory-bus fuckup with GTX 970.
Posted on Reply
#28
Dragokar
So NVIDIAs statement is busted?
Posted on Reply
#29
Zazigalka
Don't panic, Blackwell refresh will fix it :p
Posted on Reply
#31
Darc Requiem
Serious question when was the last time a GPU series launch was such a comedy of errors?
Posted on Reply
#32
neatfeatguy
At this rate, as long as AMD's cards don't burst into flames when you open the box they come in, I think AMD should do pretty good this time around.
Posted on Reply
#33
Rahmat Sofyan
"hey you want MFG ( more fakesss generated )... we give it all"

RTX 5090
RTX 5080
RTX 5070Ti

all top three cards from nvidia at least for now, has the same problem, very rare or intended problem
Posted on Reply
#34
Dr. Dro
More and more I find myself in this exact same spot. Thanks Mikel, you are a prophet.

Posted on Reply
#36
Wirko
Bruno VieiraYou know, no one will RMA them, as most of them were from scalpers, and the other ones are afraid to take six months to have the GPU back. Keep on milking guys 100% yields.
And the rest of them will be put to work that demands all the shaders and not a single ROP.
Posted on Reply
#38
TheGeekn°72
WirkoAnd the rest of them will be put to work that demands all the shaders and not a single ROP.
Unless, somehow, Nshittia managed to have dies missing shader units off the prod line as well. :laugh::laugh::laugh:
Posted on Reply
#39
Tatty_Two
Gone Fishing
Caring1Found the ONE person that uses Bing. :roll:
Bonus.......... 2 people, my youngest daughter does too, me tho, only on my tablet NOT PC :clap:
Posted on Reply
#40
_roman_
ImoutoYou know they are going to stop reporting them in the driver like they did with hotspot temps.
The performance difference is still there. People claim 11%, so I assume it's around 15%.

Some other side had technical spreadsheets for those gpus. Basically the scammend user bought a lower tier card for full price.
Bruno VieiraYou know, no one will RMA them, as most of them were from scalpers
RMA denied. RMA rules most often is only for the first buyer of the product. Not for resale.

When you bought it personally in a shop, you may pass on the receipt.
neatfeatguyAt this rate, as long as AMD's cards don't burst into flames when you open the box they come in, I think AMD should do pretty good this time around.
Like Kids in the Supermarket. I want that green Nvidia graphic card. Not the red AMD one or the blue Intel one. wahhhhhhhhhhh.

I highly doubt many nvidia buyers will not buy an nvidia card again.

Just shift the topic to AMD so we can not trash "talk" about nvidia graphic cards.

We should talk about NVIDIA, their quality department, the board partners, about fraud, about scam, about selling a product which is basically a 5060 ti, ...
Posted on Reply
#41
Lycanwolfen
I miss the old days when Nvidia and Ati only made one card each. Back in the day when the 9800 Pro came out or the Geforce 6700 one card done no other options quality control intact and it worked.
Who ever is dumb enough to pay 3000.00 for a video card is beyond me for a 5090. Just was not long ago like 2016 you could build a whole computer with video card under 1000 dollars.
Posted on Reply
#42
Tomgang
OMG a shit show Nvidia has right now. Paper launch, still melting connections on at least rtx 5090, some cards is said to die from the lastens driver with black screen and bsod and deffective gpu dies with mission rops on both 5090 and 5080. For not to forget the overpriced cards to and a heftig increase in power consumption to.

I can't think of a bigger shit show since well fermi gtx 480 any one or maybe the gtx 970 with that vram performance decrease on the last 512 mb vram or so, where memory brandwith decreased significantly.

This is a disaster launch and gpu general.

Remember gtx 970

9gag.com/gag/aZxoeOV

Or gtx 480

www.reddit.com/media?url=https%3A%2F%2Fi.redd.it%2Fyas8gg9dms111.jpg
Posted on Reply
#43
lilunxm12
Anyone know whether blackwell performance is particular sensitive to rops?

3060 (GA104) has 33% more (64 vs 48) rops, but performance seems identical to regular 3060
Posted on Reply
#44
TheGeekn°72
lilunxm12Anyone know whether blackwell performance is particular sensitive to rops?

3060 (GA104) has 33% more (64 vs 48) rops, but performance seems identical to regular 3060
We got the 5090 losing about 5% performance with the 8 missing ROPs, the 5070Ti losing 10-12% performance, also due to 8 missing ROPs and now we got reports of a 5080 missing 8 ROPs, yet to be tested for amount of missing performance.

Now, there are games that aren't necessarily ROP intensive but most will be dependent on those
Posted on Reply
#45
Ruru
Trepie
It's not a defect, it's a feature. Also an user error. -Nvidia
Posted on Reply
#46
Vya Domus
Worst thing about this most people wont even know, users on these forums are a minority, lots of people out there without a clue about any of this, they'll just randomly get a worse product and never find out.
lilunxm12Anyone know whether blackwell performance is particular sensitive to rops
Every GPU is sensitive to number of ROPs, it's one of the main limiting factors in graphics performance, it's why for instance Vega 64 could not match a 1080ti despite having similar compute capacity, 64 vs 88 ROPs.
Posted on Reply
#47
Ruru
Trepie
Vya DomusWorst thing about this most people wont even know, users on these forums are a minority, lots of people out there without a clue about any of this, they'll just randomly get a worse product and never find out.
Prebuilt owners are one group of people I can think of.
Posted on Reply
#48
Lew Zealand
Vya DomusWorst thing about this most people wont even know, users on these forums are a minority, lots of people out there without a clue about any of this, they'll just randomly get a worse product and never find out.
100%. The average user won't know they're missing 5% or even 10% performance if they plug in a 5090 to replace a 4080, or 5080 to replace a 4070/Ti/Super. It's still gonna be noticeably faster.
Vya DomusEvery GPU is sensitive to number of ROPs, it's one of the main limiting factors in graphics performance, it's why for instance Vega 64 could not match a 1080ti despite having similar compute capacity, 64 vs 88 ROPs.
It's not 1:1 though as the 3070 vs 4060 Ti is real interesting with both performing and scaling similarly but with the 3070 having 96 ROPs to the 4060 Ti's mere 48. The 4060 Ti's 26% lower core count is overcompensated by its ~45% higher clock speed but you'd think the average of those things would be kneecapped by a halving of ROPs. Yet the 4060 Ti matches the 3070. I wonder if the 4060 Ti had 64 ROPs would it be considerably faster than the 3070?
Posted on Reply
#49
Ruru
Trepie
Lew ZealandIt's not 1:1 though as the 3070 vs 4060 Ti is real interesting with both performing and scaling similarly but with the 3070 having 96 ROPs to the 4060 Ti's mere 48. The 4060 Ti's 26% lower core count is overcompensated by its ~45% higher clock speed but you'd think the average of those things would be kneecapped by a halving of ROPs. Yet the 4060 Ti matches the 3070. I wonder if the 4060 Ti had 64 ROPs would it be considerably faster than the 3070?
Wonder how much does the pure architecture improvement help between Ampere and Ada. Like IPC and other stuff beside just raw numbers compared.
Posted on Reply
#50
lilhasselhoffer
So...let me get this straight. Nvidia immediately comes up with the 0.5% figure. They immediately cite three models. It's almost like they already knew that some percentage of "good" product that they released was in fact testing badly. At that failure rate, we are looking at 5 in 1000, or one in 200. That doesn't sound like a lot...but these are thousand dollar+ cards that just have some number of ROPs that don't work. If I'm cynical, they are looking at their AI frame generation to make-up for this failure, effectively "hiding" it behind what could theoretically be silly levels of interpolated frames, meaning instead of 1 real in 6 frames their algorithm might have to generate 1 real in 5 frames...but the difference of effectively 220 fps and 180 fps would be indistinguishable on a 120 Hz refresh rate monitor.

While I won't defend the scuminess, I do think that someone somewhere along the chain knew the fallout was happening and thought it'd be a reasonable risk. I also think that the 0.5% figure is shenanigans. When you hear half a percentage you are prone to not consider it your problem. If they announced that 2% of units would experience drops then everybody would test...because that's 1 in 50 people. Given that they have no legal obligation to actually disclose failure rates expressly to anyone (including their shareholders if this is slid into their nominal silicon failure rates), then we'll never actually know how badly they tried to screw consumers.
Posted on Reply
Add your own comment
Nov 21st, 2026 08:48 PST change timezone

New Forum Posts

Popular Reviews

TPU on YouTube

Controversial News Posts