How much would you pay for RTX 4080?

[Dr. Dro]

AMD doesn't believe in ray-tracing. They said they it will be fully supported and available only in the cloud.
So, not mainstream and not for you.

View attachment 270730

You're taking this (very old, pre-RDNA2) slide quite out of context. AMD isn't going to be running raytracing server farms for Radeon owners, cloud computing is aimed at the application specific market.

And not for me? Not sure what you mean by that. I mean, I know I'm only a hobo that still has an RTX 3090 (smh I don't have a 4090 yet, what am I, poor?), but... I dunno, I enjoy raytraced gaming, even if my wood GPU only gets 100 fps or whatever, wtf how am I so poor, playing at 1080p and not using frame generation

I see people talking about cost per mm2 for production costs. Yep, sure that's increased. But did you even look to see the mm2 used by each GPU?

RTX 3080 : 628.4mm2
RTX 4080 : 379mm2

Even if their per mm costs have increased the die size has drastically decreased, by 40%. It definitely does NOT justify the massive cost increase to the cards. Nvidia are being greedy, it's a corporation after all, we expect them to do that. The problem is AMD isn't being competitive, and neither is Intel (in this high-end space). Nvidia has the market by the balls, you don't buy AMD cause they aren't very future proof, and you don't buy nvidia (but you will) cause they're too expensive.

Those people who say they don't believe in Ray Tracing, go live on an intel integrated and tell me you're still fine with it for gaming. Graphics goes forwards, Ray Tracing solves problems that typical shader based raster programs find difficult to scale, and we've been finding difficult to remedy for a decade now without dedicated hardware. Traditional triangle based rasterization is at it's limit of being efficient, and you might not think it, but taking the Ray Tracing route is about making certain effects MORE efficient, because otherwise you have to brute force them with traditional shader programs, which end up slower (grab a GTX 1080 and use it to run Quake 2 RTX, they have the entire ray tracing stack running in shaders).

The explanation for the die sizes is quite straightforward: The RTX 4080 is built on a much more advanced lithography node and is a lower segment ASIC (AD103) compared to the highest tier die (the AD102), and the RTX 3080 had a seriously cutdown GA102, only 68 out of the 84 computing units present in the GA102 are enabled in an RTX 3080. This number increases to 70 on 3080-12GB, 80 in the 3080 Ti and 82 in the 3090, with the 3090 Ti having a fully enabled processor. Using high yield harvested (low-quality!) large dies with several disabled units tends to hurt power efficiency. Add first-generation GDDR6X memory and it's no wonder the original Ampere has seen better days.

But indeed, I agree, Jensen has raised prices this generation quite a bit. But not only that, they've also left an insane amount of space for refresh SKUs, and even a comfortable lead above the RTX 4090 for an eventual 4090 Ti or potential 30th anniversary release Titan Ada or something, as the RTX 4090 has only 128 out of the 142 units of the AD102 processor enabled.

 

[Wasteland]

But indeed, I agree, Jensen has raised prices this generation quite a bit. But not only that, they've also left an insane amount of space for refresh SKUs, and even a comfortable lead above the RTX 4090 for an eventual 4090 Ti or potential 30th anniversary release Titan Ada or something, as the RTX 4090 has only 128 out of the 142 units of the AD102 processor enabled.

Right, Nvidia's left all sorts of room for segmentation at the extreme high end. What they haven't done, seemingly, is to leave room down the stack for Ada's generational uplift to shine. The 4090, at its $1600 MSRP, actually offers better performance per dollar than the 4080 at $1200. This sort of value:cost progression is unprecedented, as far as I know. The value sweet spot always resided in the low-to-mid range.

Anyway, if we extrapolate the trend downward, Ada cards in the <$700 range probably wouldn't perform noticeably better--possibly worse, even--than their Ampere counterparts (at their original MSRP), at least with regard to traditional metrics. Of course, it doesn't look like Nvidia intends to compete with its own Ampere stack any time soon.

I know Nvidia announced a "4060" to be released next Summer (lol), but barring aggressive competition from AMD, we may have to wait until the 50-series before we see a genuine attempt on Nvidia's part to craft a coherent pricing scheme from top to bottom. Until then, we can argue about costs, yields, wafer sizes, profit margins, etc, but all of that discussion is completely overwhelmed, IMO, by the lingering effects of the crypto-mining boom, during which Nvidia undoubtedly gorged itself--and ended up with a huge pile of excess Ampere cards as a result. That remains the most obvious and significant explanation for Nvidia's behavior.

 

[ARF]

You're taking this (very old, pre-RDNA2) slide quite out of context. AMD isn't going to be running raytracing server farms for Radeon owners, cloud computing is aimed at the application specific market.

You simply lack knowledge and understanding about how many transistors are needed to run ray-tracing in real-time.
The AMD Vision presentation slide explains to such as you that you will never get so much power in a single graphics card. Because you need trillions of transistors.
And this is achievable only by supercomputers which are in the cloud.

 

[Dirt Chip]

nVidia's profits rose "more sharply" than the costs you referred to did.

While it's certainly true that the costs to manufacture increased, the prices nVidia sells their cards have increased A LOT MORE, which is how nVidia manages to make such "obscene" profits.

I agree, so what?

I also disagree with protest buying, though I have to add that AMD has pretty competitive products at reasonable prices across their entire stack, so I don't believe you end up with a lesser product right now if you choose AMD for whatever reason. With AMD paving the way for the 7900 series, there are some pretty sweet deals to be had on an RDNA 2 GPU. I've just recently bought a 6750 XT for £470. The competing 3070 starts at at least £100 higher. Some models even touch the £700 mark which is insane. The 6800 and 6900 XT are heavily discounted as well, not to mention the 6600 series which has been a best buy ever since its release. The only Nvidia card I consider worth buying at current retail prices is the 3080 for £700-750. Unfortunately, it's out of my price range (although I'm happy with the 6750 XT, so I don't really need it).

Agreed, AMD have good offer nowadays, and RT is not your favorite I see no reason not to go AMD on a cost to preformance basis.

 

[AusWolf]

I see people talking about cost per mm2 for production costs. Yep, sure that's increased. But did you even look to see the mm2 used by each GPU?

RTX 3080 : 628.4mm2
RTX 4080 : 379mm2

Even if their per mm costs have increased the die size has drastically decreased, by 40%. It definitely does NOT justify the massive cost increase to the cards. Nvidia are being greedy, it's a corporation after all, we expect them to do that. The problem is AMD isn't being competitive, and neither is Intel (in this high-end space). Nvidia has the market by the balls, you don't buy AMD cause they aren't very future proof, and you don't buy nvidia (but you will) cause they're too expensive.

Those people who say they don't believe in Ray Tracing, go live on an intel integrated and tell me you're still fine with it for gaming. Graphics goes forwards, Ray Tracing solves problems that typical shader based raster programs find difficult to scale, and we've been finding difficult to remedy for a decade now without dedicated hardware. Traditional triangle based rasterization is at it's limit of being efficient, and you might not think it, but taking the Ray Tracing route is about making certain effects MORE efficient, because otherwise you have to brute force them with traditional shader programs, which end up slower (grab a GTX 1080 and use it to run Quake 2 RTX, they have the entire ray tracing stack running in shaders).

What are you talking about? A 6600 for £320 isn't competitive against a £370 3060? Or a £460 6700 XT against a £550 3070? Or a £630 6800 XT against a £740 3080 10 GB or £815 3080 12 GB? These are the cheapest prices of each tier at Scan UK. Except for the 3060, AMD offers the same performance with more VRAM (which I think is more future-proof - though I don't like this word) at lower prices. The point is: except for the $1,000+ halo market, there isn't a single tier of GPUs right now where AMD doesn't offer you more for less money.

 

[ARF]

What are you talking about? A 6600 for £320 isn't competitive against a £370 3060? Or a £460 6700 XT against a £550 3070? Or a £630 6800 XT against a £740 3080 10 GB or £815 3080 12 GB? These are the cheapest prices of each tier at Scan UK. Except for the 3060, AMD offers the same performance with more VRAM (which I think is more future-proof - though I don't like this word) at lower prices. The point is: except for the $1,000+ halo market, there isn't a single tier of GPUs right now where AMD doesn't offer you more for less money.

The current prices in Germany in euro, RX 6600 is 260:

Radeon RX 6400 - 152.89
Radeon RX 6500 XT - 199.00
Radeon RX 6600 - 259.00
Radeon RX 6600 XT - 339.00
Radeon RX 6650 XT - 329.00
Radeon RX 6700 XT - 399.00
Radeon RX 6750 XT - 455.00
Radeon RX 6800 - 499.00
Radeon RX 6800 XT - 648.00
Radeon RX 6900 XT - 699.00
Radeon RX 6950 XT - 880.00

 

[TheoneandonlyMrK]

I don't really believe in "ray-tracing". I am fine with normal lighting. And I think nvidia ngreedia is a scam, it makes fake marketing "technologies".

Meh..

Nobody Wants NVIDIA's $1199 US GeForce RTX 4080: Despite Lower Shipments, Retailers & Stores Are Stocked With Cards (wccftech.com)

You know the more I see Ray tracing.
And I can use it.

The more I think it's a complete waste of time, energy and money.

Pre baked lighting is so good in many cases on what are fast moving game's that to me the difference is not worth the effort.

Plus instead of pre baked lighting giving us efficient and effective gaming.

We're all supposed to buy cards using twice the eco credits (power) for three times the money credits, all while trying to save the earth by not using plastic bags cars or farting too much.

I think it lunacy.

 

[trog100]

it seems that the majority think a 4080 should cost 7 or 8 hundred dollars.. this figure seems to be linked to what a 3080 card was released at even though it was impossible to buy one at that price..

ether way 7 or 8 hundred dollars is not cheap its not in the average gamers ballpark.. luckily there is plenty of cheaper options available..

i was one of the people who paid £1100 quid for a 2080ti so i am not surprised at what a 4080 is launched at.. but maybe buying a 2080ti altered my way of thinking.. it must have..

trog

 

[Dr. Dro]

You simply lack knowledge and understanding about how many transistors are needed to run ray-tracing in real-time.
The AMD Vision presentation slide explains to such as you that you will never get so much power in a single graphics card. Because you need trillions of transistors.
And this is achievable only by supercomputers which are in the cloud.

Transistors have... nothing to do with it? What you need is a suitable architecture with enough compute power to do it. That slide is at least 3 years old, and from the company that at the time did not offer an RT capable architecture. So... yeah. You shouldn't assume people don't understand things when they clearly do.

 

[BSim500]

Those people who say they don't believe in Ray Tracing, go live on an intel integrated and tell me you're still fine with it for gaming.

That makes no sense. Whether Ray-Tracing exists or not doesn't mean non-RTX dGPU's would have been "uninvented". It's like saying to someone "You don't see the need for PCIe 5.0 SSD? Go use a 1.44MB floppy disk then" as if there's absolutely nothing in between of those two extremes...

 

[AusWolf]

it seems that the majority think a 4080 should cost 7 or 8 hundred dollars.. this figure seems to be linked to what a 3080 card was released at even though it was impossible to buy one at that price..

ether way 7 or 8 hundred dollars is not cheap its not in the average gamers ballpark.. luckily there is plenty of cheaper options available..

i was one of the people who paid £1100 quid for a 2080ti so i am not surprised at what a 4080 is launched at.. but maybe buying a 2080ti altered my way of thinking.. it must have..

trog

The 2080 Ti was the fastest commercially available card in the Turing lineup, while the 4080 isn't.

The other thing is that covid lockdown ebay scalper prices don't count when one talks about MSRP.

In my opinion, Nvidia should have done the same as what AMD is doing right now (and what both companies have been doing for years): massively discount their current lineup, clear all inventory, then release the new series for the same price as the last one. It's less of a cash-grab, but it's the safer option for the company's image.

 

[Dirt Chip]

I hope AMD, with the new MCM design, will be able to do to NV what they did to Intel with Zen.
But their slow gaining in the RT front might left them behind. AI and deep learning can scale preformance much better and NV dominant it without a question.
AMD doesn't develop those capabilities as fast as NV and their opening place is behind NV.

 

[64K]

The 2080 Ti was the fastest commercially available card in the Turing lineup, while the 4080 isn't.

Not sure why anyone would compare the MSRP of the 4080 to a 2080 Ti. They are 2 entirely different categories. Wait till we see the price of a 4090 Ti when/if it arrives. It's going to be ridiculously expensive from the looks of things.

The reality is that the 4080 is merely a midrange GPU that is seriously overpriced but the whole Ada stack will be as well.

 

[trog100]

Not sure why anyone would compare the MSRP of the 4080 to a 2080 Ti. They are 2 entirely different categories. Wait till we see the price of a 4090 Ti when/if it arrives. It's going to be ridiculously expensive from the looks of things.

The reality is that the 4080 is merely a midrange GPU that is seriously overpriced but the whole Ada stack will be as well.

i dont consider a 4080 mid range.. its high end.. not the highest but its for sure not mid range..

trog

 

[64K]

i dont consider a 4080 mid range.. its high end.. not the highest but its for sure not mid range..

trog

That's your prerogative. I won't bother debating it with you if your mind is made up. For everyone else compare the die size and memory bus width of the 4090 to the 4080 and it will be clear that the 4080 is a midrange Ada GPU. The 4090 is the only high end Ada for now but my guess is that there will be a 4090 Ti at some point. Links to the database here below:

NVIDIA GeForce RTX 4090 Specs

NVIDIA AD102, 2520 MHz, 16384 Cores, 512 TMUs, 176 ROPs, 24576 MB GDDR6X, 1313 MHz, 384 bit

www.techpowerup.com

NVIDIA GeForce RTX 4080 Specs

NVIDIA AD103, 2505 MHz, 9728 Cores, 304 TMUs, 112 ROPs, 16384 MB GDDR6X, 1400 MHz, 256 bit

www.techpowerup.com

 

[trog100]

i want just talking about available 4xxxx series cards i was talking about cards available for sale.. nvidia do seem to be running both 3xxx and 4 xxx series cards alongside each other.. looked at this way a 4080 card is definitely a high end card..

gone are the days when one series ended and another began.. as for how much a high end card should cost i really dont have a clue.. they cost what they cost.. you buy one or you dont.. nobody has to buy this latest generation.. the 3xxxx series cards are still plenty good enough..

trog

 

[64K]

i want just talking about available 4xxxx series cards i was talking about cards available for sale.. nvidia do seem to be running both 3xxx and 4 xxx series cards alongside each other.. looked at this way a 4080 card is definitely a high end card..

gone are the days when one series ended and another began.. as for how much a high end card should cost i really dont have a clue.. they cost what they cost.. you buy one or you dont.. nobody has to buy this latest generation.. the 3xxxx series cards are still plenty good enough..

trog

Nvidia isn't running Amperes alongside Ada. There were just too many Amperes made and they are still taking time to be sold. That happened because Nvidia was trying to meet demand which was artificially high due to miners. Mining dropped off abruptly and so there was a glut of Amperes remaining to be sold. Obviously they could have cleared the retail channel of Amperes quickly if prices had been lowered considerably.

 

[trog100]

Nvidia isn't running Amperes alongside Ada. There were just too many Amperes made and they are still taking time to be sold. That happened because Nvidia was trying to meet demand which was artificially high due to miners. Mining dropped off abruptly and so there was a glut of Amperes remaining to be sold. Obviously they could have cleared the retail channel of Amperes quickly if prices had been lowered considerably.

they are running both series side by side.. exactly why i dont know but to be honest it makes sense to me even if it dosnt to you..

trog

 

[Dirt Chip]

Not sure why anyone would compare the MSRP of the 4080 to a 2080 Ti. They are 2 entirely different categories. Wait till we see the price of a 4090 Ti when/if it arrives. It's going to be ridiculously expensive from the looks of things.

The reality is that the 4080 is merely a midrange GPU that is seriously overpriced but the whole Ada stack will be as well.

Nvidia isn't running Amperes alongside Ada. There were just too many Amperes made and they are still taking time to be sold. That happened because Nvidia was trying to meet demand which was artificially high due to miners. Mining dropped off abruptly and so there was a glut of Amperes remaining to be sold. Obviously they could have cleared the retail channel of Amperes quickly if prices had been lowered considerably.

We can debate the reasons, but they officially running them alone.
See 4080\4090 NV lunch slides.

 

[AusWolf]

they are running both series side by side.. exactly why i dont know but to be honest it makes sense to me even if it dosnt to you..

Well, maybe they shouldn't. They're just too greedy to admit that Ada was meant to be a replacement for Ampere, but it isn't due to unforeseen circumstances.

All of this could have been avoided by delaying the Ada launch altogether, but that would have put AMD in the advantage with RDNA 3. They wanted to cash in on miners, but now it's gonna bite them in the ass.

gone are the days when one series ended and another began.. as for how much a high end card should cost i really dont have a clue.. they cost what they cost.. you buy one or you dont.. nobody has to buy this latest generation.. the 3xxxx series cards are still plenty good enough..

That's not true. Just look at what AMD is doing. They're heavily discounting RDNA 2 cards to clear inventory and make space for RDNA 3. This is what both AMD and Nvidia have done during every product launch.

 

[Acesbong]

700 max. Taking people for fools, they are selling the gimmick of ray tracing to gamers so the can just use one sku for most markets.

 

[Razrback16]

Just don't go out and buy a card if you don't really need one. Both AMD and Nvidia need to realize that the pandemic is over and people won't pay bonkers prices anymore.

This is good advice - and I'd add on to say if you *do* need a card really bad, don't be afraid to look at used cards. I've found some really good deals on ebay over the last few years just being patient and knowing what I was willing to pay for them.

 

[caroline!]

Couldn't vote but $0

I'm not paying for real-time clowntracing.

 

[ARF]

Fixed it:

This tier had never had so small and slow for the tier chip

 

[evernessince]

This is good advice - and I'd add on to say if you *do* need a card really bad, don't be afraid to look at used cards. I've found some really good deals on ebay over the last few years just being patient and knowing what I was willing to pay for them.

Yes, you can find some crazy good used deals when a lot of cards are on the market. Bough 2 RX 580s for friends back in 2019 at $80 / each that are still working like new today.