NVIDIA GeForce RTX 4080 Founders Edition

[Eva01Master]

Same place he'll be able to get the other one for $900

 

[mb194dc]

Price is way too high. In a world where prior gen is averaging 80fps at 4k, who's spending the extra for this ?

I guess we'll see. Both teams cards are far too expensive.

 

[CyberCT]

Glad I picked up my EVGA 3080ti for $820 on Newegg during a ridiculous sale they had months ago. Never saw one go that low again and NVIDIA must be keeping supply low on purpose to keep demand and prices high ... even the 30 series cards went up in price, which is ridiculous.

Having a Samsung Gsync TV was the best purchase I made in a while. I don't care if the 3080ti can't push 120fps constantly at 4k ... The image quality is incredible and I don't notice the FPS fluctuation between 70-120 too much. Gonna skip this generation for sure.

 

[medi01]

If AMD's presentation figures hold, even non XTX 7900 will be somewhat slower trade blows, mostly win against 4080 16GB.

Gsync ...was the best purchase I made in a while

Ironic.

 

[Dr. Dro]

Glad I picked up my EVGA 3080ti for $820 on Newegg during a ridiculous sale they had months ago. Never saw one go that low again and NVIDIA must be keeping supply low on purpose to keep demand and prices high ... even the 30 series cards went up in price, which is ridiculous.

Having a Samsung Gsync TV was the best purchase I made in a while. I don't care if the 3080ti can't push 120fps constantly at 4k ... The image quality is incredible and I don't notice the FPS fluctuation between 70-120 too much. Gonna skip this generation for sure.

Anyone sane would do what you're doing. Ada is just a turnoff, the prices are too high, NV is as usual stingy with backporting features to former flagship cards... and honestly this architecture is more of the same. RDNA 3 on the other hand is an impressive technical achievement... AMD's design has soul this time around.

Eagerly awaiting the Navi 31 reviews, but were I a betting man, I'd stake on the 7900 XT wrecking this card and the 7900 XTX being eerily competitive with the RTX 4090 - at a much lower price and power consumption figure. We'll have to wait and see.

 

[Deleted member 185088]

Great review though as usual here!
As for the card, it's stupidly overpriced just like the 4090, and then the new RT cores what are they doing, the card loses performance similarly to Ampere, only the 4090 is step up in this regard.
Again wasted silicon.

 

[rv8000]

Insignificant performance jump over the 3090ti @4k, previous gen cards are plenty strong at 1080/1440p to drive 100+ FPS.

This product has zero value at its current price. People will flock to buy it while complaining about price, and never break the endless cycle.

 

[Fleurious]

Looks like a decent enough card but i’m tired of these ever increasing prices for xx80/xx90 cards so i’ll just stick with my trusty 1070ti until it dies.

 

[DemonicRyzen666]

If AMD's presentation figures hold, even non XTX 7900 will trade blows, mostly win against 4080 16GB.

View attachment 270069


Ironic.

most of those games list for RX 6950 XT are almost 10% higher than Techpowerup's listed frames in reviews here.

 

[N3M3515]

How can a rip off be an editors choice?

 

[Dr. Dro]

most of those games list for RX 6950 XT are almost 10% higher than Techpowerup's listed frames in reviews here.

W1zz has been getting some flak for the stock 5800X testbench setup for some time now. It's a realistic setup which I personally love but if you're looking for maximum potential, it's not gonna cut it for current gen cards. The CPU just cannot do it.

 

[skates]

These 4080 reviews are disappointing in that they are all editors choice and one of the main factors in considering cards for a lot of people is price, which the reviews only list as a con. TPU has been my go-to for many years and this is a missed opportunity to tell Nvidia that pricing counts and remove editors choice, rather than just being a con. This is the first set of reviews which I suspect are biased and I've never felt that way with TPU until now.

 

[AnotherReader]

most of those games list for RX 6950 XT are almost 10% higher than Techpowerup's listed frames in reviews here.

Probably different test scenes.

 

[mb194dc]

How can a rip off be an editors choice?

My guess would be Nvidia not happy if the conclusion is? : whilst it's a technically very impressive card, most use cases will be better served buying the prior gen at the cheapest price they can.

If you've got money to burn or want that extra 10 or 20fps at 4k + RT on, go for it...

That's the truth, as far as I'm concerned anyway.

 

[skates]

As a long time Nvidia supporter I love all the new tech and performance, but the prices are too much. There is actually a chance I might go AMD for the 1st time since the 15+ year old 5000 series. Which is annoying because DLSS 3.0 and RTX are amazing. Or I might stay with my 2080 another Gen!

Me too. I have an 2080Ti and this is the first time in ages that I've skipped a generation (3000 series) and will now skip this generation as well. I've never owned AMD, but I'll be getting the XTX on release because it has DP 2.0 and I'm interested in a DP2.0 monitor next year and secondly, I just don't want to support Nvidia any longer, their pricing and availability is ridiculous.

 

[medi01]

most of those games list for RX 6950 XT are almost 10% higher than Techpowerup's listed frames in reviews here.

Yeah, just noted. Thanks.

 

[clopezi]

5800X in test setup...again huge bottleneck and flattened results especially in 1080p/1440p.

Not huge at all, but anyway, who wants one of those cards to play 1080P games? C'mon, be realistic here.

 

[neatfeatguy]

As I'm looking over the results I'm thinking, "Damn, this is a lot faster than I though it was going to be....out pacing the 3090Ti, WTF?" Then it clicked that I was thinking about the 4080 12GB. For some reason I had the crappy, what is now defunct, 4080 12GB on the brain.

The power use is lower than I thought it would be, so I guess that's a good thing going for the card.
The performance is about where I figured it would be based on the released graphs.
The RT hasn't improved over Ampere. Ada is still suffering roughly a 40% drop in performance with RT on, same as Ampere.
The price, though, is shit. Nearly double the price of previous 3080 10GB and even if you compared it to the 12GB model, it's still a $400 premium price you're paying.

If this card came in right at $799, I could see it being a worthwhile purchase. I hope the 7900XT (or at least the XTX) comes in and walks all over this card. Come on AMD, don't let us down because If I was in the need for a high end card this gen I'd rather spend $900-1000 instead of $1200. Thankfully, though, I'm not in need for a new GPU.

 

[Vecix6]

There is a typo in the first page charts: says AD102 and 76300M transistors, but in the intro says: "NVIDIA developed a new in-between silicon between the two, to better optimize manufacturing costs, the AD103." and if I not missed this is a 45900M transistors and close to 380 mm2

 

[medi01]

It is notable that with RT (aka "kill my fps") on, perf drop is roughly the same as in the last gen.

What happened to "new GPUs will have more RT, so fps won't tank" promise?

 

[Icon Charlie]

Price is high because Nvidia wants to force potential buyers to take the 4090 instead for not much more.

IMHO.... there is more to this. I will not say much (since I'm from Silicon Valley and I hear lots of interesting things and I don't want to be one day Epstein'ed... heh), BUT IHMO the reason the pricing is this high is because....

1. Too many Greenies that Drank the cool aid for so many years that they have been zombified. So when Lord Gonzo puts on his leather Jacket and says... "Go!!!... and buy our expensive video cards so that you can be one of the cool guys and feel important and entitled!!!".... They do so with grunts and groans. The shambling hoards bang on their keyboards, buying their stuff without worrying about how much it costs. Because... They are the Greenies!!!...

2. Leather Jacket Lord Gonzo knows that his easy money from Crypto is over. He needs price it so that it looks gooooood for the share holder and his fat paycheck he gets every year.

Okay... so there you have it.... I will now hop on my Gyrocopter and fly towards my make shift shelter in the deepest and darkest parts of the technology jungle and try to hide... And with gallons of soapy water for my sprayer and cases of "Old Spice" Deodorant to hurl at the Greenies, I know they will melt and bubble away... as personal hygiene to the Greenies is like holy water to the undead. I shall defend myself against the Ngreedia Greenies to the Bitter End!!!

Soooooo.... Get Ready with your secret Decoder Ring and your Ovaltine for next week's Episode on the WKFOOL radio in:

Icon Charlie and the Secrets of Silicon Valley!!!

 

[TheinsanegamerN]

Not huge at all, but anyway, who wants one of those cards to play 1080P games? C'mon, be realistic here.

That argument has been beaten to death. Many times.

A GPU test tests the GPU. This may shock you, but that means that your test rig should not bottleneck the GPU. You are testing the theoretical output of hte card, not based on how realistic the final setup is. Because by that logic, we shouldnt test at 4k.

Same reason CPU tests include 720p results, to make the CPU test the bottleneck

 

[medi01]

So, MSRP (assuming they were real):

3080: $699 (cryptobazinga inflation made that much more than that)
4080: $1199 (current pricing I see in DE is 2199 to 2521 Euros )

Performance gain from next gen: +50%
Price bump of the next gen: +70%

Rich.

Editor's Choice.

Same reason CPU tests include 720p results, to make the CPU test the bottleneck

This was beaten to death, but I don't recall any consensus on such tests making sense (and in my opinion, it doesn't, obviously)

 

[Valantar]

Yeah, as was mentioned above here, @W1zzard, it feels like quite the stretch to give an "Editor's choice" award to a product that increases its price over its predecessor by 70%.

 

[TheinsanegamerN]

This was beaten to death, but I don't recall any consensus on such tests making sense (and in my opinion, it doesn't, obviously)

The entire argument against it is "well nobody plays like that". Which has nothing to do with the argument, which was that low rez tests in CPU testing are used to force the CPU to be the bottleneck to explore the theoretical limit of the chip. (and for GPUs, using the fastest CPU available to ensure that the GPU is the bottleneck for GPU tests, because the entire point of GPU tests is to test the GPU)

When your entire rebuttal of an argument is whataboutism it tends to get thrown away.