Next Gen GPU's will be even more expensive

[freeagent]

My crystal ball says:

Prices will increase until people stop buying. Then, and only then will prices start to come down again.

Remember, during the human malware days people showed these companies what they were really willing to pay..

 

[Lost_Troll]

Well, I do not know if anybody saw the post on Videocardz about the interview with Jack Hyunh had with Tom's Hardware.

In the interview he says the dropped the enthusiast graphic cards to focus on the lower and mid tear cards to get a better overall market share then what AMD has now. They also want to work with the developers closer than they have in the past to gain more scale in the market. He also hints that price is also reason for them getting out of the flagship GPU arena with the PS 5 price comment. I also think that Intel's GPU's are now cutting into their market share as well and they see them as a threat.

To add to this, Moors Law is Dead also said pretty much the same thing a few weeks ago and also said AMD was shooting for a mid 400$ price for the 8800XT/XTX, which would be great if it does have better raytracing then the 7900XT with the same raster rate as well.

To me, I think it is a smart move by AMD to do this. If they can focus on making lower and mid tear cards faster and cheaper, they can get a better market share. I also believe that it will allow them to have their engineering division to focus on raytracing and path tracing and make it better. Now they just need to get FRS up to par with DLSS and not make it hardware bound like Nvidia did.

 

[sepheronx]

I can barely play some games with the 2080 Ti on potato preset and the 3090 is 50% faster so it looks like I won't last a year with the 3090, not even a day. this is madness

I call bullshit on this.

The rtx 3070 runs great in most games at 1440p and the 2080ti is a better 3070 with the 11gb of vram.

Which is plausible. It will somewhat follow the ada playbook although the 4080 was a bit better vs the 3090ti than this would be vs the 4090.....

When competitor isn't moving perfomance forward this is what we get....

I figured an 1800-2000 for the 5090 due to 4090s seemingly selling ok at around that makes sense.

And all those who are still on 3000 series might see a 1200 usd 4090 + 10% likely with 16GB of vram good enough after waiting so long....

I'm on a 3000 series card.

Pray tell what games need super high end? Wukong? Yeah, wonderful, a souls rehash game. Do I care? No. The problem is, games are usually extreme shit these days and ones worth playing don't require something stupid to play it. My 3080 runs Alen wake 2 fine. It runs dying light 2 fine. It runs resident evil games fine. Games actually worth playing. I'll even give wukong a try and see how well it runs at 1440p. Can't play ultra settings? I'll go down a few notches and bet it will look just fine.

People like tpu users are the problem. Buy every generation and tell people who may not know much about PC's that "its a good time to buy as it's a reasonable upgrade from 3xxx series". Yeah, that's bullshit and reason why Nvidia can screw everyone over with these prices.

 

[Hyderz]

I mean, what more do you want from AMD/Nvidia? Halo cards are upsold on the general consensus of them being halo cards, but they aren't any different from all the rest - just a bit faster, but so what? Being the fastest available is not a virtue. Next year, the next best thing comes out, and then, that will be the fastest. There's nothing new or interesting about it.

Ppl want the best and fastest, it makes them feel good and they want to brag about. Also they feel superior when other people have shittier hardware… basically young kids mentality

 

[AusWolf]

Ppl want the best and fastest, it makes them feel good and they want to brag about. Also they feel superior when other people have shittier hardware… basically young kids mentality

And that's why graphics cards are more expensive than ever: We've got more idiots with money in the tech/gaming communities than normal people.

 

[chrcoluk]

Yeah I agree with both of those. A lot of the advances are just used for laziness (for lack of a better word - don't want to go too hard at developers I know its not easy for them right now) but thats essentially what it is. Who was it that said ' the real programming starts when you run out memory' well, computers have a lot of ram these days, and still pretty fast ssd paging files behind that. Every new version of Word or photoshop uses more resources and does mostly the same things. I remember running the first photoshop on a computer with 1MB of memory. I mean of course there are improvements (and more os overhead), but the graphs don't seem to be scaling at the same rate.

Dx12 and vulkan may be 'closer to the metal' which in theory gives room for more game optimizations, but thats assuming you have the time to actually do that. A lot of games are made with pre-made engines these days anyway. I like how that gives an easier in for beginners, but the engines have problems of their own. Shader compilation stutter on UE has been a problem for a long time. Programmers CAN do things to lessen it but may not always have time. They are always being pushed for deadlines.

And with all these tools making programming easier - still development times have ballooned. Perhaps its the size of the projects. Maybe its a factor that that so many veteran programmers that know the tricks you used to need, are retiring. There's also all the layoffs to consider (or maybe its the opposite - too many cooks the kitchen). But taking a whole generation to make a game is ridiculous and obviously unsustainable. Maybe projects need to be scaled down? idk. I think we're right in the middle of some kind of shift right now.

Just please - dont let it be more f2p games. I am so often going through steam, something catches me eye only to see the dreaded 'free to play'

Yep it may be out of control of the devs in some situations as well, there was a video I recently linked to on here, and in the video it showed a game dev open a request on the UE forum to get a performance problem with the engine fixed and Epic replied with a "wont fix" and that was that.

 

[oxrufiioxo]

Pray tell what games need super high end? Wukong? Yeah, wonderful, a souls rehash game. Do I care? No. The problem is, games are usually extreme shit these days and ones worth playing don't require something stupid to play it. My 3080 runs Alen wake 2 fine. It runs dying light 2 fine. It runs resident evil games fine. Games actually worth playing. I'll even give wukong a try and see how well it runs at 1440p. Can't play ultra settings? I'll go down a few notches and bet it will look just fine.

People like tpu users are the problem. Buy every generation and tell people who may not know much about PC's that "its a good time to buy as it's a reasonable upgrade from 3xxx series". Yeah, that's bullshit and reason why Nvidia can screw everyone over with these prices.

Only you can decide what's worth spending your money on nobody should dictate when ir if you need an upgrade, that doesn't mean we shouldn't expect perfomance to be improved every two years.

No gpu can be considered a necessity unless you are making money from it and paying your bills with it how much your hobby is worth to you as far as gaming is your own business and nobody's else.

I decide what my hobbies are worth based on my disposable income if that is more or less than someone's else's it doesn't matter or at least shouldn't matter.

What I buy or don't buy has 0 bearing on what others should or shouldn't buy.

I mean some people own nicer houses than other people doesn't make having a roof over your head any less important.

People buy what they can afford if person A can easily afford 1600 but person B can only afford 300 it doesn't make the hobby more or less important to either person.

I have a budget system with a 6700XT a mid tier one with a 3080ti and a high end one with a 4090 gaming Is awesome on all 3 that doesn't mean I shouldn't want or expect somthing better with the next generation at all 3 levels.

 

[evernessince]

Biggest is.... Rt is easier for developers

I have yet to see anyone prove this argument.

RT is only easier in the sense that to get the same level of graphic fidelity, it requires fewer tricks (generally). Developers still have to place individual lights in their levels and set lighting properties just the same as if they were using rasterized lighting. Mind you, UE5 tends to close the gap between rasterized and ray traced lighting. Software based lumen is run without dedicated RT cores and the quality is quite good. Devs get that right out of the box with that engine so no extra effort is needed beyond what you'd have to do regardless of whether you want your game to support RT or not.

A lot of people seem to be forgetting that devs have to implement RT on top of their rasterized lighting, so without a doubt it's only adding to how much work they have to do. Hence why most implementations are Nvidia sponsored. In addition, from the games I've seen thus far with RT, the vast majority of them are designed with rasterization first in mind and RT is tacked on. I have yet to see a dev that goes back through the game and hand tunes every light in every scene, probably because that would be extremely time consuming and mostly a waste of time for the few users that would benefit. Assuming that RT will eventually become more popular, you are now essentially requiring devs to ensure their games look good with two entirely different lighting systems. That's a massive amount of work.

To top it off, RT requires heavy optimization and there is far far ecosystem support and documentation for rasterized lights as compared to RT. Imagine how much work devs put into optimizing their current game lighting and then consider that RT increases the performance hit by a minimum of 3x up to 50x+ (RT will get more demanding as we move away from a mere 0.5 samples per pixel and the current 0 bounces and each small step will massively increase the performance hit).

Realtime RT has got to be the holy grail of selling video cards for the next 15 years at least, particularly if prices keep going the way they are. It makes it so easy to push new video cards, a simple upping of the sample rate from 0.5 to 1 will nuke performance on the 4000 series or heck adding 1 more RT light source to a scene or even allowing a single bounce.

My crystal ball says:

Prices will increase until people stop buying. Then, and only then will prices start to come down again.

Remember, during the human malware days people showed these companies what they were really willing to pay..

Doesn't apply today. Nvidia has the healthcare , enterprise, DPU, and AI markets to fall back on. It would be extra-ordinarily hard to get enough games to make Nvidia take notice.

 

[chrcoluk]

Well it was an actual dev that told me RT is great for devs, they just place a light source and the hardware handles the rest. The same dev told me legacy lighting was all done manually, like your reflection in the water wasnt live, it was something they had to do manually to get the effect, so in terms of development resources RT is apparently a massive win. It probably does add extra optimisation requirements, but optimisation is usually only a small part of game dev anyway, as its the lowest priority of development.

 

[AusWolf]

Only you can decide what's worth spending your money on nobody should dictate when ir if you need an upgrade, that doesn't mean we shouldn't expect perfomance to be improved every two years.

No gpu can be considered a necessity unless you are making money from it and paying your bills with it how much your hobby is worth to you as far as gaming is your own business and nobody's else.

I decide what my hobbies are worth based on my disposable income if that is more or less than someone's else's it doesn't matter or at least shouldn't matter.

What I buy or don't buy has 0 bearing on what others should or shouldn't buy.

I mean some people own nicer houses than other people doesn't make having a roof over your head any less important.

People buy what they can afford if person A can easily afford 1600 but person B can only afford 300 it doesn't make the hobby more or less important to either person.

I have a budget system with a 6700XT a mid tier one with a 3080ti and a high end one with a 4090 gaming Is awesome on all 3 that doesn't mean I shouldn't want or expect somthing better with the next generation at all 3 levels.

I would agree, but there's another way to look at it. If there's no improvement, then there's no reason to upgrade, so you can spend your disposable income on other things. You win either way.

 

[sepheronx]

I would agree, but there's another way to look at it. If there's no improvement, then there's no reason to upgrade, so you can spend your disposable income on other things. You win either way.

The problem I have is that people not in the know will respect someone like his opinion because of the years he's been here and contributions. So people will take to heart what he suggests to them. Essentially waste money. While I think he is a good contributor for sure, I strongly disagree with his position.

The reason why we are in the spot we are in with this overpriced hardware is because people are paying this stupid prices. And the more we recommend people to upgrade when upgrade makes no sense, is contributing to this very problem.

Whatever.

 

[AusWolf]

The problem I have is that people not in the know will respect someone like his opinion because of the years he's been here and contributions. So people will take to heart what he suggests to them. Essentially waste money. While I think he is a good contributor for sure, I strongly disagree with his position.

The reason why we are in the spot we are in with this overpriced hardware is because people are paying this stupid prices. And the more we recommend people to upgrade when upgrade makes no sense, is contributing to this very problem.

Whatever.

A fair point.

In my opinion, for any upgrade to make financial sense, one has to factor in the cost of the upgrade (new GPU buying price - old GPU selling price) compared to the improvement in performance. If the performance improvement is bigger than the spending, then that's a good upgrade, otherwise, it's not, and should not be recommended. Naturally, this equation goes out of the window if one considers themselves an "enthusiast", and upgrades only because of why not, and not because it makes sense financially.

 

[iameatingjam]

I have yet to see anyone prove this argument.

RT is only easier in the sense that to get the same level of graphic fidelity, it requires fewer tricks (generally). Developers still have to place individual lights in their levels and set lighting properties just the same as if they were using rasterized lighting. Mind you, UE5 tends to close the gap between rasterized and ray traced lighting. Software based lumen is run without dedicated RT cores and the quality is quite good. Devs get that right out of the box with that engine so no extra effort is needed beyond what you'd have to do regardless of whether you want your game to support RT or not.



A lot of people seem to be forgetting that devs have to implement RT on top of their rasterized lighting, so without a doubt it's only adding to how much work they have to do. Hence why most implementations are Nvidia sponsored. In addition, from the games I've seen thus far with RT, the vast majority of them are designed with rasterization first in mind and RT is tacked on. I have yet to see a dev that goes back through the game and hand tunes every light in every scene, probably because that would be extremely time consuming and mostly a waste of time for the few users that would benefit. Assuming that RT will eventually become more popular, you are now essentially requiring devs to ensure their games look good with two entirely different lighting systems. That's a massive amount of work.

I sure hope by a lot of people you don't mean me, because that was kinda the whole point of my post. As to to why I think games will slowly move towards rt only, once that mid tier hardware is capable of it..... That way game developers don't have to do the lighting twice, and you can do it the easier way. (or I guess you contest that now, or do you? seems like you do, then kind of cast doubt on it in the next paragraph). Of course by easy I didn't mean 0 work is required, I meant less work is required. You have to place light sources yes, but you don't have to do nearly as much work with reflections and shadows.

We've already seen an rt only game with Star Wars outlaws.

To top it off, RT requires heavy optimization and there is far far ecosystem support and documentation for rasterized lights as compared to RT. Imagine how much work devs put into optimizing their current game lighting and then consider that RT increases the performance hit by a minimum of 3x up to 50x+ (RT will get more demanding as we move away from a mere 0.5 samples per pixel and the current 0 bounces and each small step will massively increase the performance hit).

Right... which is why in most games right now you can turn RT off. What we were talking about is where we think game development is heading. Not where it is right now. I don't even like this direction, but nonetheless I think its probably where things are going, just due to, what I said before, capitalism, its easier to implement and companies like to simplify their operations. I think there will be software advancements and hardware advancements and together, we'll get mainstream raytracing for everybody. It wont be fancy pants maxed out raytracing, but raytracing nontheless.

But I can't predict the future. Only giving my perspective on what I think will happen. I've always been perfectly happy with our fake rasterized lighting. We got pretty damn good at it too. And it obviously depends on the game but I think in a lot of recent titles it has taken a back seat to ray tracing, otherwise... our damn mirrors would work!

Doesn't apply today. Nvidia has the healthcare , enterprise, DPU, and AI markets to fall back on. It would be extra-ordinarily hard to get enough games to make Nvidia take notice.

nvidia has a lot of other sources of revenue yes, but does that mean they would willingly give up the gaming market, just cause they have something else? I don't think so, not unless things got really dire and suddenly drastic cuts were needed. The rules of supply and demand still apply to them.

 

[evernessince]

Well it was an actual dev that told me RT is great for devs, they just place a light source and the hardware handles the rest. The same dev told me legacy lighting was all done manually, like your reflection in the water wasnt live, it was something they had to do manually to get the effect, so in terms of development resources RT is apparently a massive win. It probably does add extra optimisation requirements, but optimisation is usually only a small part of game dev anyway, as its the lowest priority of development.

Placing lighting, RT or otherwise isn't as simple as just placing the light. You still have decide the light type (spot, directional, point, ect), light color, and the many other light properties. Certain lights will disperse photons in a specific direction with a very sharp cutoff while others will disperse them around the lighting fixture. More realistic light propagation simulation ala RT doesn't change these requirements. You are still going to want to adjust these parameters regardless of whether you are using RT or rasterized lighting. In addition, some devs may want to bake lighting when fully dynamic lighting isn't required and other times they will need to use different types of reflections (aside from SSR, UE5 supports cheaper reflection captures) because having a ton of RT reflections isn't remotely feasible. Devs are going to use a gambit of reflection types according to their requirements including RT, SSR, and reflection captures (which includes planar, box, ect).

You may have misunderstood what they said, implementing RT lighting over rasterized lighting might be easy because you've already done most of the work setting the correct lighting properties but implementing RT lighting does not absolove you of doing that work in the first place. In addition you may still have to go through and adjust lighting parameters for ideal visibility under RT. I've noticed in many RT enabled games the devs don't do this and often text or objects that are supposed to be visible are harder to see. If the dev you talked to really did think that you can simply place a generic light source throughout their entire game without editing it's properties then I very much doubt they were much of a dev to begin with.

I sure hope by a lot of people you don't mean me, because that was kinda the whole point of my post. As to to why I think games will slowly move towards rt only, once that mid tier hardware is capable of it..... That way game developers don't have to do the lighting twice, and you can do it the easier way. (or I guess you contest that now, or do you? seems like you do, then kind of cast doubt on it in the next paragraph). Of course by easy I didn't mean 0 work is required, I meant less work is required. You have to place light sources yes, but you don't have to do nearly as much work with reflections and shadows.

We've already seen an rt only game with Star Wars outlaws.

Star Wars outlaws having RT on by default doesn't mean the game doesn't use rasterization (it does, hence why you get ok performance on cards like the 1080 Ti). It also uses hybrid RT, so a portion of the work is done on tradition shader cores.

I'm sure we'll move to RT only at some point but people have been saying "once mid tier hardware is capable of it" since the 2000 series (heck there were even some bold individuals claiming mid tier 2000 series cards were it already) and here we are, mid-tier GPUs now costing $700 USD and requiring upscaling just to run modern games with RT enabled at a decent frame-rate. The progress has been less than impressive.

It's not even the first time I've heard someone say 'nice mid-tier hardware is capable of RT well get RT only games'. That was 6 years ago now and we are still waiting. I'd MUCH rather have ray-traced sound, much less resource intensive and extremely beneficial. Even better, dynamically generate sounds based on how sound propogation work. Every pot should make a different sound based on the weapon hitting it, where it hits it, ect. Literally anything other than graphics for once. Decent AI for example sure would be nice. Modern AAA games are as wide as an ocean and as shallow as a puddle.

 

[oxrufiioxo]

I would agree, but there's another way to look at it. If there's no improvement, then there's no reason to upgrade, so you can spend your disposable income on other things. You win either way.

Yes winning at life is much better than winning the next GPU upgrade. Doesn't mean people should hate progress because it gives them less reason to upgrade... A 6700XT isn't gonna perfom any worse if a 5090 is 2x a 4090....

Right now the market just sucks it sucks for everyone but I'd argue it's worse for the sub 500 usd market than it's ever been. Even as a high end gpu buyer that bothers me the most we can't have progress if the entry level sucks multiple generations in a row.

 

[iameatingjam]

Star Wars outlaws having RT on by default doesn't mean the game doesn't use rasterization (it does, hence why you get ok performance on cards like the 1080 Ti). It also uses hybrid RT, so a portion of the work is done on tradition shader cores.

I'm sure we'll move to RT only at some point but people have been saying "once mid tier hardware is capable of it" since the 2000 series (heck there were even some bold individuals claiming mid tier 2000 series cards were it already) and here we are, mid-tier GPUs now costing $700 USD and requiring upscaling just to run modern games with RT enabled at a decent frame-rate. The progress has been less than impressive.

It's not even the first time I've heard someone say 'nice mid-tier hardware is capable of RT well get RT only games'. That was 6 years ago now and we are still waiting. I'd MUCH rather have ray-traced sound, much less resource intensive and extremely beneficial. Even better, dynamically generate sounds based on how sound propogation work. Every pot should make a different sound based on the weapon hitting it, where it hits it, ect. Literally anything other than graphics for once. Decent AI for example sure would be nice. Modern AAA games are as wide as an ocean and as shallow as a puddle.

Thats not why it works on the 1080 ti, it has a software fallback, it doesn't require RT accelerators to work, but its still ray tracing. Have you seen it with RT force disabled? It doesn't look like that on the 1080 ti.

Though both do look pretty bad...

 

[oxrufiioxo]

Placing lighting, RT or otherwise isn't as simple as just placing the light. You still have decide the light type (spot, directional, point, ect), light color, and the many other light properties. Certain lights will disperse photons in a specific direction with a very sharp cutoff while others will disperse them around the lighting fixture. More realistic light propagation simulation ala RT doesn't change these requirements. You are still going to want to adjust these parameters regardless of whether you are using RT or rasterized lighting. In addition, some devs may want to bake lighting when fully dynamic lighting isn't required and other times they will need to use different types of reflections (aside from SSR, UE5 supports cheaper reflection captures) because having a ton of RT reflections isn't remotely feasible. Devs are going to use a gambit of reflection types according to their requirements including RT, SSR, and reflection captures (which includes planar, box, ect).

You may have misunderstood what they said, implementing RT lighting over rasterized lighting might be easy because you've already done most of the work setting the correct lighting properties but implementing RT lighting does not absolove you of doing that work in the first place. In addition you may still have to go through and adjust lighting parameters for ideal visibility under RT. I've noticed in many RT enabled games the devs don't do this and often text or objects that are supposed to be visible are harder to see. If the dev you talked to really did think that you can simply place a generic light source throughout their entire game without editing it's properties then I very much doubt they were much of a dev to begin with.



Star Wars outlaws having RT on by default doesn't mean the game doesn't use rasterization (it does, hence why you get ok performance on cards like the 1080 Ti). It also uses hybrid RT, so a portion of the work is done on tradition shader cores.

I'm sure we'll move to RT only at some point but people have been saying "once mid tier hardware is capable of it" since the 2000 series (heck there were even some bold individuals claiming mid tier 2000 series cards were it already) and here we are, mid-tier GPUs now costing $700 USD and requiring upscaling just to run modern games with RT enabled at a decent frame-rate. The progress has been less than impressive.

It's not even the first time I've heard someone say 'nice mid-tier hardware is capable of RT well get RT only games'. That was 6 years ago now and we are still waiting. I'd MUCH rather have ray-traced sound, much less resource intensive and extremely beneficial. Even better, dynamically generate sounds based on how sound propogation work. Every pot should make a different sound based on the weapon hitting it, where it hits it, ect. Literally anything other than graphics for once. Decent AI for example sure would be nice. Modern AAA games are as wide as an ocean and as shallow as a puddle.

Until 4090 is the entry level we likely won't see AAA games developed with it in mind and we also really won't know have much larger a visual gap a game can have until its developed from the ground up with RT in mind with no fallbacks. My guess is 2-3 generations.

Although with how entry level cards are so neglected maybe 4 generations.

Right now even with the path traced games we are in the tact on after the fact generation becuase that's all the hardware can run.

The RT game I'm most looking forward to currently is old AF Half Life 2 and beyond that Witcher 4. I think the cadence is going to get slower and slower regardless of almost everyone using UE5.

The PS6/Xbox next if there is a nextbox will likely dictate how much a jump we actually get in the short term though.

People shouldn't worry though there will always be AA and indie game that don't have RT so them toasters will keep chugging along and the way it's going the majority of good game will fall in this category anyways.

 

[AusWolf]

Yes winning at life is much better than winning the next GPU upgrade. Doesn't mean people should hate progress because it gives them less reason to upgrade... A 6700XT isn't gonna perfom any worse if a 5090 is 2x a 4090....

I don't hate progress. I just don't get stressed out when there is none.

Right now the market just sucks it sucks for everyone but I'd argue it's worse for the sub 500 usd market than it's ever been. Even as a high end gpu buyer that bothers me the most we can't have progress if the entry level sucks multiple generations in a row.

That's very true. Personally, I don't give a rat's arse about a 5090 level GPU, and I probably never will. One can call it "halo", or "enthusiast" or just "more money than brains", it's still just a damn toy.

 

[oxrufiioxo]

I don't hate progress. I just don't get stressed out when there is none.


That's very true. Personally, I don't give a rat's arse about a 5090 level GPU, and I probably never will. One can call it "halo", or "enthusiast" or just "more money than brains", it's still just a damn toy.

It's mostly annoying as I hate stagnation but that applies to a lot of things for me not just gpu's.

I don't lose any sleep over it though.

If everyone gave up making desktop gpus tomorrow I'd happily game on a console but I expect the same, that there is progress each generation....

I actually game on both PS5 and Switch and plan on getting a PS6 and Switch 2.

My hobby is still gaming with the gpu just a small but very expensive portion of that lol. It also isn't my most expensive hobby. Nvidia is trying their hardest to change that though lol....

 

[iameatingjam]

It's not even the first time I've heard someone say 'nice mid-tier hardware is capable of RT well get RT only games'. That was 6 years ago now and we are still waiting. I'd MUCH rather have ray-traced sound, much less resource intensive and extremely beneficial. Even better, dynamically generate sounds based on how sound propogation work. Every pot should make a different sound based on the weapon hitting it, where it hits it, ect. Literally anything other than graphics for once. Decent AI for example sure would be nice. Modern AAA games are as wide as an ocean and as shallow as a puddle.

Well I actually hope you're right about that one. I don't like tracing and I avoid using it to save power. All I was saying is, I think this is the direction the AAA games are going. Like I said I can't predict the future. Its just my educated guess. Moreso that software raytracing will be simplified and optimized than mid tier hardware will get powerful enough to do what we consider ray tracing today, but it will probably be a bit of both.

 

[oxrufiioxo]

Well I actually hope you're right about that one. I don't like tracing and I avoid using it to save power. All I was saying is, I think this is the direction the AAA games are going. Like I said I can't predict the future. Its just my educated guess. Moreso that software raytracing will be simplified and optimized than mid tier hardware will get powerful enough to do what we consider ray tracing today, but it will probably be a bit of both.

In 10 years it'll be funny looking back at this conversation.... when a 60 class card starts at 6-700 usd and we miss the good old days of ada.

We didn't know how good we had it

 

[iameatingjam]

In 10 years it'll be funny looking back at this conversation.... when a 60 class card starts at 6-700 usd and we miss the good old days of ada.

We didn't know how good we had it

And I'm sure 60 class will be a generous word for it.

Though if consoles still dominate 10 years from now, this will probably be happening on amd. And those able to afford high end nvidia cards can crank it up even more. Then again 10 years is a long time. who know what can change. Maybe we'll have diferrent players, maybe China will invade taiwain and tsmc will get blown up. Can you imagine the gpu drought when that happens?

Maybe the US facility comes to fruition, maybe it doesn't.... Samsung may get some new customers though.

 

[SL2]

I think it could be a smart move for AMD to stay at mid range for now, even if some customers will disapprove. Or move.. maybe they don't have a choice.

AMD could maybe make something faster than a 7900 XTX right now, but the question is if it would sell well enough. I'd guess no, it might not be worthwhile for them, especially in times of inflation.

It looks like they're now about to do what they did with the 5700 XT or even the HD 3870 (yup, 17 years ago) which is not trying to/(not being able to) chase the fastest card from the competition. (Yeah, they could do it for various reasons unrelated to the competition, or whatever reason I don't know anything about lol)

Make the products attractive and build demand before adding halo models, and if the former doesn't happen.. try again.

AMD currently has one RX 7000 model (7800 at #13) in amzn's top 25 sales list, that's not good enough. Then we have the RX 6000 cards still hanging around..

I want my medal now lol

AMD strategy shifts to prioritize market scale over enthusiast Radeon GPUs - VideoCardz.com

Jack Hyunh reveals AMD’s plan to build market Share, not chase flagship GPUs The company will have something for enthusiasts eventually. Jack Hyunh, AMD’s senior vice president and general manager of the Computing and Graphics Business Group, was interviewed by Tom’s Hardware at IFA 2024, where...

videocardz.com

Edit: And now I saw the TPU post

 

[InVasMani]

With AMD out of the high end market and essentially leaving Nvidia the $600+ segment just for themselves and apparently Nvidia using the more expensive GDDR7 memory, as well as decent increase in cores on a more expensive process node as well, the next gen high end graphics are going to be even more expensive, so if you were hoping for the next generation to bring in some sanity, think again!

Latest leaks show the 5090 will used around 600W, over 20k cores and will cost around $2500. The next GPU down the line the 5080 will cost around $1500 and only the mid range 5070 non TI to cost bellow $1000 or around $800 dollars, with the 5070TI costing $1000.

Assuming AMD is very competitive in the mid range segment with aggressive pricing, then we can expect the 5060TI to cost around $500, the 5060 around $400 and the 5500 around $300.

AMD are likely to undercut Nvidia in the mid and low end segment with prices of the 8800XT from $500, 8700XT $400, 8600XT $300 and 8500XT $200.

The new generation needs to be previous generation a tier above it on price, performance, and efficiency by about 5-10% in each of the three categories ideally. It really doesn't matter what AMD is doing that's what Nvidia should be aiming to do with coming generation of hardware. AMD is focusing a bit more on lower end and mid range and value more heavily than in the past, but it shouldn't change much with how Nvidia approaches things.

The only reason Nvidia users care is they want AMD to drive down the cost of Nvidia hardware and ramp up more competitive innovation at the high end of the market. What AMD's doing should drive more competition though in the low end and mid range which I'd say is where the emphasis should be to make newer hardware technology more accessible.

Nvidia can raise prices all it wants, many cards are already priced at what the market will bear.

People were already extending how long they keep their cards to help accommodate for increased pricing but $1,000 for a 5070 Ti that will have a tiny memory bus and little VRAM doesn't sound anywhere remotely appealing.

People think AMD was keeping Nvidia's pricing in check but AMD and Nvidia have not really competed on pricing since polaris. After that AMD has simply been pricing around Nvidia. Nvidia currently has 88% of the market, for all intents and purposes the market is less competitive by the numbers than the 10 year period where Intel had a monopoly of the CPU market. It's worse by a significant margin to boot, AMD held 26% of the CPU market with bulldozer but it currently holds a mere 12% of the GPU market. By extension current GPU pricing already has Nvidia's monopoly pricing baked in.

Nvidia did benefit from Intel's anti competitive behavior against AMD. That's partly how we got to this point in the first place. On top of that Nvidia had it's own GameWorks controversy. I think most people are recognizing that Nvidia has a bit too much of a competitive edge over the competition right now to the point it's bad for consumers and businesses alike and really only benefits for investors. One company having a commanding competitive lead in over 50% in any market is generally not a great thing. Obviously things aren't going to be perfectly balanced, but it's far out of balance now.

 

[iameatingjam]

I want my medal now lol

AMD strategy shifts to prioritize market scale over enthusiast Radeon GPUs - VideoCardz.com

Jack Hyunh reveals AMD’s plan to build market Share, not chase flagship GPUs The company will have something for enthusiasts eventually. Jack Hyunh, AMD’s senior vice president and general manager of the Computing and Graphics Business Group, was interviewed by Tom’s Hardware at IFA 2024, where...

videocardz.com

Edit: And now I saw the TPU post

After hearing the reasoning behind it, I agree. Most people just want an affordable, good value gpu. It will be an uphill battle though, and the gpu will have to be a way better value than nvidia for most people to cross over, not just a bit better. Can amd deliver that? Idk.