- My Forums
- Tiger Rant
- LSU Recruiting
- SEC Rant
- Saints Talk
- Pelicans Talk
- More Sports Board
- Fantasy Sports
- Golf Board
- Soccer Board
- O-T Lounge
- Tech Board
- Home/Garden Board
- Outdoor Board
- Health/Fitness Board
- Movie/TV Board
- Book Board
- Music Board
- Political Talk
- Money Talk
- Fark Board
- Gaming Board
- Travel Board
- Food/Drink Board
- Ticket Exchange
- TD Help Board
Customize My Forums- View All Forums
- Show Left Links
- Topic Sort Options
- Trending Topics
- Recent Topics
- Active Topics
Started By
Message
re: PC Discussion - Gaming, Performance and Enthusiasts
Posted on 10/6/23 at 6:26 am to UltimateHog
Posted on 10/6/23 at 6:26 am to UltimateHog
I'm all for competition, but Epic went about it the wrong way. I will always oppose companies paying to exclude competitors. Hopefully they can refocus on offering something to the market that Steam doesn't instead of lazily brute forcing the issue.
1
Posted on 10/6/23 at 1:08 pm to Joshjrn
quote:
I sincerely hope the reviews for FSR3 are great. Nvidia needs competition in that arena.
*sigh*
HWU did a deep dive on the two available games with FSR 3: LINK
There are a fair number of issues with how the tech works right now that I won't go into (watch the video if you're interested) because AMD has already acknowledged the problems and says they are working on it.
Beyond that, it looks like the frame gen tech is good enough. It's not as good as Nvidia's, but it's fine. Though there are some caveats that come along with that that are too detailed to go into, but are covered well in the video. The problem is that "FSR 3" is really just "FSR 2.2 with frame gen". There has been zero improvement to the core upscaling technology, which currently lags painfully behind DLSS in visual quality, and you can't use AMD's frame gen tech without also using their upscaling tech.
Actually did a bit of googling, and while the video is much more illustrative, the conclusion at the bottom of this article basically covers the issues in a bit more depth than I have: LINK
I hope the press this is getting is a wake up call for AMD, but I have to think Jensen is cackling right now.
Posted on 10/6/23 at 1:20 pm to Joshjrn
Yeah frick Epic. This was always going to be the outcome.
Posted on 10/6/23 at 6:54 pm to UltimateHog
DF FSR3 Hands On
Pretty much the consensus is the image quality is really good (he says he is still quite taken aback by how good this software solution is and up there with the competition), has a few bugs as expected at first launch and DF reached out to AMD and they confirmed VRR is already working with FSR3 for next the update.
I love how unbiased DF always is, alluding to the poor shape and many issues DLSS3 had at launch.
Pretty much the consensus is the image quality is really good (he says he is still quite taken aback by how good this software solution is and up there with the competition), has a few bugs as expected at first launch and DF reached out to AMD and they confirmed VRR is already working with FSR3 for next the update.
I love how unbiased DF always is, alluding to the poor shape and many issues DLSS3 had at launch.
This post was edited on 10/6/23 at 7:00 pm
Posted on 10/6/23 at 8:15 pm to UltimateHog
Watched the video, and his takeaway was essentially the same as everyone else: lots of bugs, which isn’t a big deal and AMD should be able to work it out. The frame gen itself is solid, though not quite as good as Nvidia. But AMD’s upscaling tech just isn’t as good as Nvidia’s, and that’s what’s causing the pronounced artifacting that he was seeing.
Frame gen needs very high frame rate to work properly. In most cases, to get a frame rate that high, you need to use upscaling. As long as FSR upscaling has pronounced ghosting and artifacting, FSR3 will be limited in its adoption.
Frame gen needs very high frame rate to work properly. In most cases, to get a frame rate that high, you need to use upscaling. As long as FSR upscaling has pronounced ghosting and artifacting, FSR3 will be limited in its adoption.
Posted on 10/6/23 at 8:32 pm to Joshjrn
quote:
FSR3 will be limited in its adoption.
This is one of the dumber things you have said.
Posted on 10/6/23 at 8:38 pm to UltimateHog
Honest question: which games have you played using FSR?
Posted on 10/6/23 at 8:44 pm to Joshjrn
What does that have to do with understanding adoption?
Availability matters the most. The fact that it's available in so many gpus new and old will all but guarantee fast adoption.
In fact, I'd bet a good sum of money a year later that FSR3 is in more games than the 25 a year into DLSS3.
Availability matters the most. The fact that it's available in so many gpus new and old will all but guarantee fast adoption.
In fact, I'd bet a good sum of money a year later that FSR3 is in more games than the 25 a year into DLSS3.
Posted on 10/6/23 at 8:57 pm to UltimateHog
quote:
What does that have to do with understanding adoption? Availability matters the most. The fact that it's available in so many gpus new and old will all but guarantee fast adoption. In fact, I'd bet a good sum of money a year later that FSR3 is in more games than the 25 a year into DLSS3.
The answer is zero, isn’t it?
And you understand that the majority of gamers are down on Nvidia frame gen as a mediocre solution in search of a problem, right? It’s not a popular feature, because its usage case is very small.
When people think high frame rate, they think smooth image and low latency. With frame gen, unless you already start at a high frame rate to begin with, you’ll likely have image jutter in spite of the high reported frame count. On top of that, you have increased latency and increased artifacting. It just doesn’t accomplish what Nvidia, and now AMD, are trying to claim it does. Most people should not use frame gen tech, regardless of whether or not they have access to it. The drawbacks aren’t worth the benefits.
But anyway, that was just a bit of general info for anyone who doesn’t immediately skip over our normal bullshite. Feel free to keep spouting your hyper biased, uninformed nonsense, though.
Posted on 10/6/23 at 9:23 pm to Joshjrn
That's what I thought.
I'll always be here to check you on your dumb comments.
I'll always be here to check you on your dumb comments.
This post was edited on 10/6/23 at 9:24 pm
Posted on 10/7/23 at 4:52 am to UltimateHog
You routinely ignore the bulk of posts that you have zero response to just to pick out single sentences to strawman. You’re a fricking clown
Posted on 10/7/23 at 5:35 am to Joshjrn
As a little bit of elaboration before I go back to ignoring you: no one on this board, other than you, gives a shite about who is “winning” between AMD, Nvidia, and Intel. They care about the tech being offered, whether it’s useful to them, and whether it’s a good price for the level of performance. When I say that FSR upscaling being subpar will “limit the adoption” of FSR frame gen, I mean that literally, and in a vacuum. I’m not saying that it will be less used than Nvidia frame gen, because I don’t care either way. It doesn’t matter to anyone other than the company’s shareholders. Well, and you, apparently.
As a gamer, I was hoping AMD would have a good showing with frame gen and improve upscaling. They had a decent showing with frame gen and did nothing to improve upscaling. Not inherently problematic, but then they tied frame gen to their upscaling tech. So someone like me, who doesn’t have access to DLSS3 might have considered using FSR frame gen in certain situations. But, now I won’t, because I would have to use FSR upscaling, which isn’t a very good experience. A personal microcosm of limiting adoption from what it could be.
But anyway, go ahead and pick your one sentence to quote and strawman. Combined with your now tacitly admitted lack of personal experience with what is discussed and your propensity for confirmation bias, I’m sure everyone will find it to be just so useful.
As a gamer, I was hoping AMD would have a good showing with frame gen and improve upscaling. They had a decent showing with frame gen and did nothing to improve upscaling. Not inherently problematic, but then they tied frame gen to their upscaling tech. So someone like me, who doesn’t have access to DLSS3 might have considered using FSR frame gen in certain situations. But, now I won’t, because I would have to use FSR upscaling, which isn’t a very good experience. A personal microcosm of limiting adoption from what it could be.
But anyway, go ahead and pick your one sentence to quote and strawman. Combined with your now tacitly admitted lack of personal experience with what is discussed and your propensity for confirmation bias, I’m sure everyone will find it to be just so useful.
Posted on 10/7/23 at 8:03 am to Joshjrn
Back to our regular non bitchy programing:
About six months ago, I posted this:
This morning, looks like we've got some rumors of a maybe:
NVIDIA reportedly working on GeForce RTX 4080 Ti with AD102 GPU for early 2024 release
Now, this could obviously be bullshite. But, if a 4080ti releases with essentially the same die as the 4090 but just with less VRAM, it becomes a very interesting buy. Though, with one caveat: When I said $1,200, I had in mind the standard late Q3/early Q4 2023 release cycle. With a Q1 2024 release, $1,200 becomes a bit harder to stomach. The 7900xtx can routinely be had for about $900 these days, and that should only see downward pressure in the next six months. Assuming the xtx stays at $900, or even manages to sink to $850 or less, a 4080ti will have a hard time justifying a price premium of $300+ for a mild to moderate raster performance bump, even with the more significant RT bump and the better upscaling technology.
At $1,400, it would be DOA for me. At $1,200, it would be tempting, but still a tough purchase for a lot of people knowing that next gen cards would coming down the pike reasonably soon. At $1,100, I think it becomes a solid buy. At $999, I think it becomes a 2020-esque "sold out everywhere, being scapled" kind of release.
But, I have approximately zero faith that Nvidia won't step on its own dick, as always, so I would say $1,200 with the 4080 being pushed to $999 is probably the best case scenario, assuming this isn't pure rumor to begin with, that it wouldn't see a significantly cut down die, etc, etc.
But, still a fun rumor for a Saturday morning.
About six months ago, I posted this:
quote:
I think there are two possibilities left for me this Gen:
1. Nvidia releases a 4080ti with a full fat 102 die but with the same 16gb of vram for $1,200. With the original 4080 sliding down to $999.
2. AMD releases a groundbreaking card with 3D cache, etc, that resets the market.
Outside of those two options, I’m likely a skip.
This morning, looks like we've got some rumors of a maybe:
NVIDIA reportedly working on GeForce RTX 4080 Ti with AD102 GPU for early 2024 release
Now, this could obviously be bullshite. But, if a 4080ti releases with essentially the same die as the 4090 but just with less VRAM, it becomes a very interesting buy. Though, with one caveat: When I said $1,200, I had in mind the standard late Q3/early Q4 2023 release cycle. With a Q1 2024 release, $1,200 becomes a bit harder to stomach. The 7900xtx can routinely be had for about $900 these days, and that should only see downward pressure in the next six months. Assuming the xtx stays at $900, or even manages to sink to $850 or less, a 4080ti will have a hard time justifying a price premium of $300+ for a mild to moderate raster performance bump, even with the more significant RT bump and the better upscaling technology.
At $1,400, it would be DOA for me. At $1,200, it would be tempting, but still a tough purchase for a lot of people knowing that next gen cards would coming down the pike reasonably soon. At $1,100, I think it becomes a solid buy. At $999, I think it becomes a 2020-esque "sold out everywhere, being scapled" kind of release.
But, I have approximately zero faith that Nvidia won't step on its own dick, as always, so I would say $1,200 with the 4080 being pushed to $999 is probably the best case scenario, assuming this isn't pure rumor to begin with, that it wouldn't see a significantly cut down die, etc, etc.
But, still a fun rumor for a Saturday morning.
Posted on 10/7/23 at 8:29 am to Joshjrn
The XTX will go down to 799 or lower if Nvidia does that.
AMD actually has a pretty good argument for the midrange now, but they aren’t even trying to compete with Nvidias high end cards.
AMD actually has a pretty good argument for the midrange now, but they aren’t even trying to compete with Nvidias high end cards.
Posted on 10/7/23 at 8:43 am to boXerrumble
quote:
The XTX will go down to 799 or lower if Nvidia does that.
AMD actually has a pretty good argument for the midrange now, but they aren’t even trying to compete with Nvidias high end cards.
That's actually something I thought about after I posted, but didn't think it was worth an edit. Both a $999 4080 and a $1,199 4080ti would apply pressure on a $900 7900xtx. I still think the xtx would be a better value for a lot of people, but for those of us who play super intensive single player games, usually with RT and upscaling, the price premium might be palatable enough to push the Nvidia offerings over the edge. If the xtx pushed down to $799, I think it's firmly in the driver's seat again. If Nvidia pushes down to, say, $899/$1,099, I'm not sure AMD can push prices low enough to stem the tide. If Nvidia blows it out and goes $799/$999, the market will explode.
But I'm not allowing myself to consider that scenario and set myself up for disappointment
Posted on 10/7/23 at 10:28 am to Joshjrn
Hmm, seems unlikely. The prices of gpus would crash. As said above, the 7900xtx goes to the 4070ti price. Then the 7900xt falls in a bad spot, it would have to drop to the 4070 price. The 7800xt would have to fall to 400 and the 7700xt to 300. This would be a blessing to AMD. You generally don't wont to throw your competitors a bone.
I mean, I hope it does, but this would snowball for Nvidia in a bad way.
The limited fab from TSMC is a bottleneck. Consumer GPU customers are tertiary from AI and enterprise.
I mean, I hope it does, but this would snowball for Nvidia in a bad way.
The limited fab from TSMC is a bottleneck. Consumer GPU customers are tertiary from AI and enterprise.
Posted on 10/7/23 at 10:35 am to hoojy
quote:
Hmm, seems unlikely.
The 4080ti in general, or potential pricing paradigms?
Posted on 10/7/23 at 10:40 am to Joshjrn
Both. It would be a weird place. At the 4080 price, you have to lower the price of the 4080, or it won't sell.
That creates a snowball effect.
I mean, I hope it does happen.
GPU prices have been insane for the last few years. This kind of snowball affect would be nice. But considering the limited capacity of TSMC? I don't see it.
At least not until late 2024 or something. Both companies will want to unload unsold gpus.
That creates a snowball effect.
I mean, I hope it does happen.
GPU prices have been insane for the last few years. This kind of snowball affect would be nice. But considering the limited capacity of TSMC? I don't see it.
At least not until late 2024 or something. Both companies will want to unload unsold gpus.
Posted on 10/7/23 at 10:56 am to Joshjrn
quote:
FSR3 will be limited in its adoption.
Nope. With how many GPU's it works with. Just nope.
Is it inferior, yes. AMD is trying a software solution for a hardware problem. With the rumors of the next RDNA being midrange, this should tell you that AMD has a shotgun ready to take RDNA out back.
Their effort is admirable, and it's decent, but software can't overcome hardware in this. Or well, at what's proven by now.
Of course, AI is the main driver in all of this.
This post was edited on 10/7/23 at 10:58 am
Posted on 10/7/23 at 12:38 pm to hoojy
quote:
Nope. With how many GPU's it works with. Just nope.
Is it inferior, yes. AMD is trying a software solution for a hardware problem. With the rumors of the next RDNA being midrange, this should tell you that AMD has a shotgun ready to take RDNA out back.
Their effort is admirable, and it's decent, but software can't overcome hardware in this. Or well, at what's proven by now.
Of course, AI is the main driver in all of this.
To go back, I don't mean it won't see any adoption. I'm saying it will be limited from what it could be. Frame gen is already a niche application. When you combine that with it being tied to FSR upscaling, I imagine tons of older Nvidia owners are just going to write it off, just like I am.
So in short, if there is a substantial number of Nvidia owners who say "meh, I'd rather have DLSS 2.2 than FSR 3 and frame gen", that's not exactly the win I think AMD was going for.
Page 1865 of 1912
Popular
Back to top
Follow TigerDroppings for LSU Football News