Reports of Bricked NVIDIA GeForce RTX 5090 and RTX 5090D Surge

[kapone32]

Wrt hardware, if it was that bad, you'd see people stop buying, class-action law suits. When you make hundreds of thousand or millions of any product, there will be defects. The defect rates are probably in the low single digits. We just have more channels to make noise about that, that's all.
And keep in mind defect rate goes up naturally as the products get more and more complex.

On a paper launch?

 

[Vayra86]

Back in the day even sli had better scaling than this monolithic core density scaling. lol.

Naeeeehh don't think so

 

[mb194dc]

So expensive and crap reliability then? Maybe why they didn't launch in volume..?

 

[dartuil]

those defective cards will do good 5070 and 5060.
joke.

 

[calhau]

And people say the 50xx series isn't that much faster.

Nvidia drivers used to take 2 to 4 years to brick their GPUs now it only takes 2 to 4 days! That's 36500% faster.

On a more serious note, now I believe der8auer, this kind of thing isn't strange if board partners only have 2 to 15 days to test the card.
And this might not be something new considering that at least the last 3 gens had major issues at release.
Never buy a GPU at release, specially a Nvidia one, seems as true as ever.

 

[Random_User]

Doesn't matter, if this is a 5600XT, or 5090. The products that being launched, should be ironed out before the release. There's no exuse. Neither nVidia, nor AMD, or Intel are garbage cheap brands and compainies. They have hundreds of billions, to find staff, that is able to fix the problems.

Once upon a time, the launch and release was meaning the same thing. Now, the "Launch" may happen month before an actual release, and yet the products often come with severe issues.

Yes, the tech has become enormously more complex. But this is still not excuse, for the misleading marketing and unlrealistic goals. If they need more time to fix the issues, the companies must keep it shut behind the doors, and not anounce anything, until the product is at least 95% ready. Otherwise, they all sell the unrealistic expectations, and increase the hype, to prepare the "hight" sales, and inflate the product and stock prices, beforehand.

Making the $2000 product, that has problems, is a slap on the face. And this isn't the first time. I mean, if someone releases such an expecive piece of tech, it should be tested before, and not upon arival, no? Maybe these people, who were waiting in queues for days (much like Aplle iPhone launch), will respect themself more, next time.

 

[truerock]

Probably more like a $3,000-5,000 brick... That would really suck

So, Nvidia will refund them $2,000?

I am joking

 

[Dr. Dro]

Doesn't matter, if this is a 5600XT, or 5090. The products that being launched, should be ironed out before the release. There's no exuse. Neither nVidia, nor AMD, or Intel are garbage cheap brands and compainies. They have hundreds of billions, to find staff, that is able to fix the problems.

Once upon a time, the launch and release was meaning the same thing. Now, the "Launch" may happen month before an actual release, and yet the products often come with severe issues.

Yes, the tech has become enormously more complex. But this is still not excuse, for the misleading marketing and unlrealistic goals. If they need more time to fix the issues, the companies must keep it shut behind the doors, and not anounce anything, until the product is at least 95% ready. Otherwise, they all sell the unrealistic expectations, and increase the hype, to prepare the "hight" sales, and inflate the product and stock prices, beforehand.

I believe it's all down to the now-widely accepted culture of using customers as beta testers.

So, Nvidia will refund them $2,000?

Since when are we taking speculation and potential problems as an irrefutable statement of fact? And of course not.

Should the need for a recall arise, it's up to stores, distributors and AIBs to refund and/or replace the defective products. But it's unlikely it will get that far.

 

[Visible Noise]

Just registered here to post.

Big snip…

Either I was just really unlucky today or this driver has some real issues.

Counterpoint: I’ve been gaming on the 570 driver branch since before it was publicly released. Only issue I have is current gen games look so good I can’t go to my backlog.

 

[Dawora]

paper launch and it's a $2k silicon brick. To those who went through all the troubles just to get their hands on a card that gives ~10% improvements over the 4090; congratulations, you've played yourself. The scalpers who bought a pallet worth of it; congratulations, you've bought a pile of bricks that no one will be buying.

10%?
im gaming 5K and its around 45% faster for those 3 games i was playing so dont know where u got this 10%?

 

[Godrilla]

Naeeeehh don't think so

lol

 

[chrcoluk]

It's not only the 5090. Roman had issues on his 5080, and trust him of all the people to know a thing or two and not be using a cheap motherboard. Dude even tried to troubleshoot it. Quickly pointing the fingers at motherboard manufacturers isn't the best thing to do tbh. He's not the only reviewer having issues either..

A very disappointing launch really, I expected more from them. It's probably the most disappointing launch i've witnessed in a good while.

Edit: Here's your chance on a silver platter AMD. You ramped production at the same time at 50 series but didn't have a joke of a launch with hundreds of products. Iron out bugs, do a proper launch, price it well and actually have stock. Judging by their track record, something should be amiss but let's not hope it's all of those points.

I wonder if that sample had an issue with the internal PCIe riser type cable the FE cards now use?

 

[mkppo]

I wonder if that sample had an issue with the internal PCIe riser type cable the FE cards now use?

It's a possibility for sure but not confirmed. At any rate, signal integrity issues happening with a riser cable is an order of magnitude greater than the issue being on the motherboard.

View attachment 383241
lol

Damn man this chart has some really good games in there. A few had even better predecessors though, Deus Ex: Human Revolution being one of the primary culprits. The original Deus Ex is still one the best games I ever played. A remake would be great, 22 years is a long time.

 

[-Mendoza-]

paper launch and it's a $2k silicon brick. To those who went through all the troubles just to get their hands on a card that gives ~10% improvements over the 4090; congratulations, you've played yourself. The scalpers who bought a pallet worth of it; congratulations, you've bought a pile of bricks that no one will be buying.

you might have to learn to educate yourself before saying anything the 10% gross gain is in 1080p... and an RTX 5090 is not bought to play in 1080p but in 4k to limit the bottleneck as much as possible.
and unfortunately for you the gain in 4k between the 4090 and the 5090 is 39% in 4k and in gross performance.

 

[Steevo]

Nvidia has hundreds of cards for sale, how dare anyone discount their work when hundreds of their cards have issues!!!


Welcome to being a hardware tester.

 

[-Mendoza-]

Nvidia has hundreds of cards for sale, how dare anyone discount their work when hundreds of their cards have issues!!!


Welcome to being a hardware tester.

I didn't understand who you were talking about? I don't have a 5000 series

 

[Vayra86]

View attachment 383240View attachment 383241
lol

What are you even saying, look closely at the bottom chart. The 5090 is +30% shaders and gets near +40% perf over a 4090, so that's more than 100% scaling. Then you show a chart of two 680's scaling for less than 70% in the vast majority of games except Starcraft, and even Starcraft doesn't achieve 100% scaling.

I think your cognitive skillset needs work, or you still need coffee

10%?
im gaming 5K and its around 45% faster for those 3 games i was playing so dont know where u got this 10%?

Yeah this topic seems to give people big issues with percentages lol

 

[529th]

i just want to know what edwardo snowden has to say about this, lol

 

[Visible Noise]

i just want to know what edwardo snowden has to say about this, lol

Somebody needs to ask Taylor Swift.

 

[529th]

Somebody needs to ask Taylor Swift.

Yeah! See what she says! PLEASE

lol

 

[lepudruk]

Well, it's a "Brickwall" after all..

 

[b1k3rdude]

Dropping the slot to PCIe gen 4 fixes it. In short, overblown, nothing to see here, move along

100%

 

[Godrilla]

What are you even saying, look closely at the bottom chart. The 5090 is +30% shaders and gets near +40% perf over a 4090, so that's more than 100% scaling. Then you show a chart of two 680's scaling for less than 70% in the vast majority of games except Starcraft, and even Starcraft doesn't achieve 100% scaling.

I think your cognitive skillset needs work, or you still need coffee


Yeah this topic seems to give people big issues with percentages lol

I guess if you compared cores from different generations sure, but if we compared the 5080 to 5090, 4080 to 4090 and the sli performance 680 to 690 the performance was almost linear. Also while you state the core density on the 5090 is 33% higher the bandwidth is unnecessary doubled which is also not stopping people from further oc the over killed vram.
I would request vram speed scaling but it seems Nvidia gave everyone 5 minutes to test the cards so there goes that. I would test it myself but it's a vapor launch rumored to be available in September.
The sli performance in the past scaled perfectly and now a chip with double the cores scales at best to 50%.
Besides the scaling what I realized Nvidia paper launched this by purpose with all the effort that went in to the Blackwell gaming cards launch. Nvidia wanted its competition to perceive that they are still focused on gaming.
Circling back with the cherry on top the proof is in the article that they aren't focused on the gaming cards with no supply and drivers that allegedly brick cards. Is it cognitive or hyper cognitive enough for you?
Now I take a hit of that caffeine.

 

[stahlhart]

I would not be surprised when some buyers of the cards changed the firmware of the card before these issues happened.



I think this is unrelated to the topic. It does not matter. It is just less electrical pcie lanes. I prefer to write electrical lanes. With the eyes I always see 16 mechanical pcie lanes PEG slots or 1 mechanical pcie slots. For the device it is important the electrical lanes.
These pcie lanes can be used for m2 nvme slots but could be used for something else or just be wasted. Recently we had another mainboard topic where some lanes were just wasted when installing a single m2 device.

--

Anyway baby burn - baby burn - with 575Watts or even more.

I would be not surprised when some of the buyers used certain software to use the card with 700 or 800 Watts to get these issues.

Some things never change.

 

[Caring1]

Well, it's a "Brickwall" after all..

I think asking Pink Floyd is more appropriate for all these bricks in the wall.