• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Closed.

Status
Not open for further replies.
I haven't had a chance to watch the presser yet, are any of the new software technologies they announced backwards compatible with the 4000 series or do they all require the 5000?
 
grifter_66 said:
I haven't had a chance to watch the presser yet, are any of the new software technologies they announced backwards compatible with the 4000 series or do they all require the 5000?
Click to expand...
Yes.

1736275500392.png
 
Hershy said:
LOL

Yup… how dare we even suggest that price will be important to decision making.
Click to expand...
All I said was that it would factor into my decision making. I never attempted to discuss the prices of the cards the way you bitched and whined about how pricing should have been part of the discussion from the beginning. That's the key difference between my point and yours. Nuance is lost on you isn't it?

The OP later amended the rules of the discussion. So now this thread is like the rest of them. Knowing the prices now and the spec differences between the 5080 and the 5090, I'm going with the 5090.
 
I'm very interested to see how the new "transformer" model for SR and RR will look. While I think they currently look great and and are quite useable, they are certainly not perfect with SR still having ghosting at times and RR still having that "boiling" look to it.
 
While on the subject of pricing, I'm typically the kind of guy who doesn't care about the price of hardware. However, in the past there have been occasions where XYZ part was 90-95% as fast as the one that was above it in the product stack but cost significantly less. In those cases I sometimes opt for the more cost effective part. So long as there aren't significant downsides. For example, I care about ray tracing performance which is why I don't buy AMD GPU's even when they aren't that much slower than their NVIDIA counterparts. That and NVIDIA drivers are generally better.

In this case I think the 5080 and 5090 are way too far apart in terms of specifications. I doubt this will be a case where the 5080 is 90% as good as a 5090. The price difference is massive and its not as if price/performance with GPU's (or anything else for that matter) is linear. I don't expect the 5080 to be half the performance of the 5090 despite being half the price. I've had the 3090 since the first weeks of it being on the market and frankly, I've got my money's worth. I don't mind going with the more expensive 5090. I'll skip the refresh (Likely a 5080Ti or 5090Ti) and then re-evaluate whether or not the 60 series will be worth getting.

I don't often skip entire generations, (normally just the refreshes) but I do on occasion depending on what they bring to the table or what else is going on in my life that might prioritize the funds going somewhere else.
 
Blackstone said:
So now the heat gets dumped up into my CPU cooler? I may need to rethink my case.
Click to expand...
It looks like the fans draw air from their side and push it through the vapor chamber and vent out the "top" .. I think?

rtx5090_fan1.jpg

rtx5090_fan2.jpg

rtx5090_fan3.jpg


and odd choice imo ... maybe time to reorient the gpu with a pci-e riser cable, push it out into no man's land. heh.
 
Last edited:
Dan_D said:
While on the subject of pricing, I'm typically the kind of guy who doesn't care about the price of hardware. However, in the past there have been occasions where XYZ part was 90-95% as fast as the one that was above it in the product stack but cost significantly less. In those cases I sometimes opt for the more cost effective part. So long as there aren't significant downsides. For example, I care about ray tracing performance which is why I don't buy AMD GPU's even when they aren't that much slower than their NVIDIA counterparts. That and NVIDIA drivers are generally better.

In this case I think the 5080 and 5090 are way too far apart in terms of specifications. I doubt this will be a case where the 5080 is 90% as good as a 5090. The price difference is massive and its not as if price/performance with GPU's (or anything else for that matter) is linear. I don't expect the 5080 to be half the performance of the 5090 despite being half the price. I've had the 3090 since the first weeks of it being on the market and frankly, I've got my money's worth. I don't mind going with the more expensive 5090. I'll skip the refresh (Likely a 5080Ti or 5090Ti) and then re-evaluate whether or not the 60 series will be worth getting.

I don't often skip entire generations, (normally just the refreshes) but I do on occasion depending on what they bring to the table or what else is going on in my life that might prioritize the funds going somewhere else.
Click to expand...

I'll be really curious to see the benchmarks, because the 5090 looks to be literally double on most things that count compared to the 5080 (CUDA core count, VRAM, memory bandwidth, and yes, price). This probably won't translate into exactly double the performance, but it will be interesting to see where exactly everything falls into place given the disparity on the stat sheet.

I don't think I've ever seen this big of a delta between tier 1 and tier 2 cards on offer from Nvidia at the consumer level. Nvidia's clearly figured out that people will pay up for the best, and they're putting a big enough gap between the best and second best to entice people to just pony up and go all in. It's an age-old retailers trick. The 4090 showed them price is less of a factor than was once thought, so the sky's the limit now as far as I can tell. This pleases me as a shareholder, but it displeases me as a consumer lol.
 
revenant said:
It looks like the fans draw air from their side and push it through the vapor chamber and vent out the "top" .. I think?

View attachment 702329
View attachment 702330
View attachment 702331

and odd choice imo ... maybe time to reorient the gpu with a pci-e riser cable, push it out into no man's land. heh.
Click to expand...

I wouldn't be overly concerned about that in a well ventilated case with a good CPU cooler. You're probably not running your CPU at a temperature that's close enough to throttle limits for it to matter. Most GPUs are already directing the hot air upwards as it is, anyway, albeit most of them also won't pump out the same amount of heat as this beast will.

Side note, as a mechanical engineer, I absolutely love the look and design of this new cooler, although I'll reserve my opinion until I see the thermal benchmarks. If it cools well though, what a fantastic piece of engineering!
 
Last edited:
HockeyJon said:
I wouldn't be overly concerned about that in a well ventilated case with a good CPU cooler. You're probably not running your CPU at a temperature that's close enough to throttle limits for it to matter. Most GPUs are already directing the hot air upwards as it is, anyway, albeit most of them also won't pump out the same amount of heat as this beast will.

Side note, as a mechanical engineer, I absolutely love the look and design of this new cooler, although I'll reserve my opinion until I see the thermal benchmarks. If it cools well though, what a fantastic piece of engineering!
Click to expand...
yeah with a well ventilated case with a lot of cross-flow should be ok.. it is a very tight little package for sure..

edit: in my current setup I'd be a little concerned about my ram as it's very close (less than an inch) to where the FE would be exhausting heat.. but I am not going with an FE so not going to fret.
 
Last edited:
Dan_D said:
All I said was that it would factor into my decision making. I never attempted to discuss the prices of the cards the way you bitched and whined about how pricing should have been part of the discussion from the beginning. That's the key difference between my point and yours. Nuance is lost on you isn't it?

The OP later amended the rules of the discussion. So now this thread is like the rest of them. Knowing the prices now and the spec differences between the 5080 and the 5090, I'm going with the 5090.
Click to expand...
Gaslighting!

All I said was that price was fundamental to the discussion.

My God…
 
Last edited:
Hershy said:
Just looking purely at the silicon there’s not much reason to upgrade from a 4090 for that price.
Click to expand...
ikr? .. the pcb looks about the size as my old evga 1050ti ... with a lot more cooling.
 
KazeoHin said:
The days of a product like the 1080Ti that can play games for generations is over. Nvidia learned to incorporate recent and relevant obsolescence in their gaming products so they don't get another product like that pesky 1080Ti that made gamers wait more than two generations to upgrade. And also, the 5080 is probably going to be $1500
Click to expand...
This x 1000.

Jensen and Ngreedia even acknowledge that the 1080 Ti is "The ultimate Geforce.", the fact that you can still run many modern games at acceptable frame rates at 1080P and even 1440P is a testament to it's longevity.

https://wccftech.com/nvidia-geforce-gtx-1080-ti-giveaway/

There is no way Jensen wants to release another legendary cost effective GPU like this which was capable of lasting 4 generations. Better to ensure constant upgrade cycles via intended obsolescence, e.g. limiting ram upgrades on any model below the halo tier.

As much as it pains me to reward Ngreedia, I pretty much have no options other than the 5090 as I need to power a 7K 240hz monitor, but at least it will be a noticeable upgrade from my 1080 Ti - maybe I should mount the 1080 on my study wall now that Nvidia has confirmed it is the GOAT.
 
Blackstone said:
Here is a great video and it discusses the lower ghosting in DLSS 4. Looks like even the 5080 is eating Cyberpunk for lunch. It looks like image quality has been significantly improved. This is a must watch!
Click to expand...
Seems like a good upgrade in image quality with DLSS 4 and updated ray reconstruction. Though issues still persist, I still saw some smearing but it was a bit less. It has 4x compute cost compared to CNN, but it's probably worth the tradeoff. It will be interesting to see how it evolves over the next few years and if they can optimize it a bit to reduce the compute cost.
 
Blade-Runner said:
Until recently raytracing has been a shitty buzz feature that was barely noticeable and not worth the performance hit.
Click to expand...
It was always noticeable when fully implemented. Most games only use one or two ray tracing features which aren't that hard on the GPU but also don't do much visually. Cyberpunk 2077, even without path tracing was extremely noticeable and came with a massive performance penalty. When the game came out, the RTX 2080 Ti was the most powerful card on the market for a short while until the 30 series came out.
 
I think it's safe to say the 5090 will only be a big upgrade in terms of RT and DLSS. So all the pure raster DLSS haters have nothing worthy to purchase and if they all choose to skip out on the 5090 like they claim they will then the rest of us shouldn't have any issues picking one up right? Riiiight??
 
MistaSparkul said:
I think it's safe to say the 5090 will only be a big upgrade in terms of RT and DLSS. So all the pure raster DLSS haters have nothing worthy to purchase and if they all choose to skip out on the 5090 like they claim they will then the rest of us shouldn't have any issues picking one up right? Riiiight??
Click to expand...
I mean you realize raster is super important still? I think the issue is, is Nvidia trying to promote a card as being this fast.....when in reality it isn't.

From Nvidia's own website the 5090 is going to be 35-45% faster than the 4090. The only strong point is DLSS 4.0. Which a game has to support......otherwise you need to rely on...guess what? Raster performance lol
 
Brackle said:
I mean you realize raster is super important still? I think the issue is, is Nvidia trying to promote a card as being this fast.....when in reality it isn't.

From Nvidia's own website the 5090 is going to be 35-45% faster than the 4090. The only strong point is DLSS 4.0. Which a game has to support......otherwise you need to rely on...guess what? Raster performance lol
Click to expand...

Yeah and the number of games that came out the past year that I played and didn't have any DLSS support I could count on one hand, not really too worried about that.
 
KickAssCop said:
For 4K 240 it doesn’t matter whether card is 20% or 50% faster as it needs to be faster. Once I hit 4K 240 in games I play, maybe I can think about taking a breather from top tier. Until that point I will use all tricks to try to hit that goal and 5090 offers flexibility to use all types of tricks to get there. I am also excited about multi frame DLSS as long as it has a quality setting that doesn’t suck.
Click to expand...

You are using a 120Hz tv, but need 240 fps?
 
Blade-Runner said:
This x 1000.

Jensen and Ngreedia even acknowledge that the 1080 Ti is "The ultimate Geforce.", the fact that you can still run many modern games at acceptable frame rates at 1080P and even 1440P is a testament to it's longevity.
Click to expand...

I don't think they even fully gave the 1080 series its full due since the 2000 series was ready with raytracing. Always felt like the silicon had more legs to run on that we had (non modded) access to.
In supported games I run an ultrawide 5760 X 1200 resolution and it handles it pretty well still. But there are a lot of games I cant keep pushing with it on reasonable settings so I need to upgrade :(
 
Blade-Runner said:
Until recently raytracing has been a shitty buzz feature that was barely noticeable and not worth the performance hit.
Click to expand...
But this describes virtually every new feature that Nvidia has introduced in the last 25 years, starting with T&L. It’s the accumulation of the features that matters. Right?

The only ‘new feature’ I’ve ever seen, with regards to gaming, that was clearly noticeable… as in, you’d have to be blind not to see the profound difference… was my first time looking at an HDR monitor with an HDR game running. The difference was night and day, with zero gpu cost.
 
Hershy said:
But this describes virtually every new feature that Nvidia has introduced in the last 25 years, starting with T&L. It’s the accumulation of the features that matters. Right?

The only ‘new feature’ I’ve ever seen, with regards to gaming, that was clearly noticeable… as in, you’d have to be blind not to see the profound difference… was my first time looking at an HDR monitor with an HDR game running. The difference was night and day, with zero gpu cost.
Click to expand...
It really depends on the game and the implementation of HDR. Not every monitor does it well either. For some people, it often looks worse than running with HDR turned off.
 
Dan_D said:
It really depends on the game and the implementation of HDR. Not every monitor does it well either. For some people, it often looks worse than running with HDR turned off.
Click to expand...
I have one HDR (OLED) monitor and one OLED handheld we use for gaming. There are a handful of games it makes a difference on, but most of the time you can't tell.
 
Blackstone said:
Brackle has valid points. The Far Cry and Plague Tale datapoints on their chart do demonstrate the relative lack of improvement in raster over 4090.

When you look at it that way, the card is somewhat disappointing—if you are already on 4090. DLSS is so ubiquitous though that it might not matter to most users. Like, even niche games like DCS World use DLSS, for example.
Click to expand...
The benchmarks for those two games were eyebrow raising. Clearly Nvidia is banking heavily on developers utilizing its tech for future releases, but I need to see multiple benchmarks for contemporary titles before I sell my 4090 and buy a 5090. At present, there’s not nearly enough information for me to drop my pants and bend over.
 
so the charts nvidia posted were 4090 dlss 3 vs 5090 dlss 4. What about where is the comparison of both cards using dlss 4?
 
I like both DLSS and FG but FG is pointless if you can't push 60FPS without it. Considering we are only seeing minor gains in raster, I'm concerned that I won't be able to maintain 60FPS in new titles since cyberpunk struggles to maintain 60FPS without FG on the 5090. It's nice to be able to flip a switch and say wow 250FPS! But it's about 5FPS away from being a laggy mess so you cut FG and drop down from 250FPS to 55FPS. Any future game that pushes harder than cyberpunk will crumble the 5090. If they can push 80 to 100 real FPS then FG is cool, otherwise it's just a clarity filter that I can't use due to low performance of my $2000 card.

I need more performance especially for non gaming related hobbies and I do think the real performance of the 5090 is concerning especially at that price point.
 
Last edited:
StoleMyOwnCar said:
That's the thing, and I don't think you're going to like it: I think that a majority of their focus this time (and probably going forward) will be AI featureset, not brute force. In other words they're saying "screw it, we can't keep up with game visual fidelity demands at all just by raw hardware needs. We're just going to increase AI tech till it gets to the point where it's perfect enough." And that might just be where the future of gaming is going (or rather, has been going for quite a while).

Although it's kind of a weird idea, isn't it? Developers keep making higher and higher fidelity games with higher and higher fidelity assets, only for those assets to get reinterpreted and re-imagined (loosely speaking) via the AI engines on modern GPUs, just so that the game can actually be playable at some level of visual fidelity within that region. At some point it will come down to exactly what I said a long time ago, the middleman will be cut out and games and their textures can be generated entirely via prompts. Soon we'll have completely procedurally generated experiences. Anyway that's a tangent and a scary thought.

It also explains why just raw RT growth numbers look like crap, but the generative AI performance is like 2x. That's basically where all of their R&D went. Nvidia is an AI first company, but instead of abandoning gaming, they simply figured out how to make AI and gaming align better.
Click to expand...
You're right, I don't like it. If they want to follow the AI fad that's all well and good, but these cards can have compute performance independent of games "requiring" DLSS and the like to have viable performance. The problem is multifaceted and a lot of it comes from engine creators (notably Epic with Unreal Engine) and game developers that are not optimizing for performance because they can "brute force" anything by tossing it under DLSS / FSR upscaling and frame generation. In the past, nobody would dare to release a title where a game is nigh unplayable on high settings without these features, and it only seems to be getting more prevalent. As far as visual fidelity, I don't think that's the case either. Up until literally this year, the vast majority of "raytracing" improvements had minor at best visual improvement for a huge performance penalty;more a parlor trick than anything else. Its possible to replicate many similar effects with baked lighting for instance, at much lower performance cost, but it takes time and effort; why bother when you can check a box on UE5 and tell people to enable DLSS? Last generation the whole justification for the 4090 was pretty much that it was the one card that you could play things with all the shiny turned on at 4K without needing resort to upscaling or other tech the way you would with just about everything else, so to see them asking more money and basically admitting that even with the increased power you're still going to need and in fact focus upon DLSS is...frustrating to say the least.

I agre that this is great for Nvidia deciding they're going to be an AI company, and their now 4-generation long marketing attempt to make RT + DLSS necessary to give it any sort of performance the "only features that matter when gaming".. but this seems to come at a growing cost to gaming, openness, and user experience overall. Nvidia deciding they want to focus on AI and compute as the latest cash cow is well enough, but selling the new paradigm to the whole industry that means less actual work/rendering being done and more reliance on upscaling and interpolated frame generation algorithms, particularly their ways of doing these things, is going to negatively affect the industry. The more that native res maximum features are no longer considerered "the benchmarking standard" by way of them being virtually unplayable, then it falls right into NV getting to defacto set the standard with DLSS and other tech like FSR will not be rated holistically valuing its openness and feature set (for instance, how NV 3000 series users' only chance at frame generation comes from AMD FSR 3.x+ variation, as NV has locked ouit those features on anything older than a 4000 series. Thus, you'd have even NV users who may use DLSS for the upscaling if they choose, but benefit from and in fact make FSR frame gen their only option for that tech) , but instead always competing with NV on their own hardware, with their own tech, with "home turf advantage"; great for Nvidia , but not so great for everyone else.

This is not an issue that is just because of NVidia. - of course they are culpable for their behavior, but it would not be nearly as much of an issue if others- from engine and game developers to reviewers - would approach this from another angle and reject the world that NV is building which favors their shareholders while everyone else is shunted into accepting a narrow, proprietary path to defining a "high end gaming experience" that costs more, incorporates planned obselescence and has other potential issues.
 
Well atleast nvidia was honest and said a while ago that future gpu performance will be mostly determined by AI techniques.
 

View: https://www.youtube.com/watch?v=4WMwRlTdaZw

if any one is thinking of water cooling DO NOT get FE. its made up 3 PCB with ribbon cable between them

Firefox_Screenshot_2025-01-09T05-45-37.272Z.png


you can see the connectors here on the back of main board that connect to PCIe (larger one on the bottom) and then to the display outs (smaller above and center).
 
I'm still confused about power cables and what exactly I would need. My current GPU is a Vetroo GV 1000 which is ATX 3.0 but in a 12+4 config not the newer 6+2. Would I still be able to use it?

I sent Vetroo an email and they said newer cables won't be available until at least April.
 
Blackstone said:
Any thoughts on Founders Edition v. third-party? I don't really really need the smaller form factor but I feel like it would make case a little easier to work inside, route cables, ect. Anyone have their eye on a specific version of 5090?
Click to expand...

The FE looks great but I'm skeptical it will be able to cool 575 watts with just 2 slots, at least while being silent. The 3-4 slot AIB coolers will be what I'm aiming for since I just upgraded my case to a North XL Mesh to support larger GPUs and when it comes to those AIB cards what I've found out is that beggars can't be choosers, you really just have to settle for whatever is available because these things are going to sell out fast.
 
Status
Not open for further replies.
Back
Top