Just looked for a potential successor for my 3080 and compared 5070, 5070 Ti and 9070XT because they have roughly the same performance. A 9070XT is about 650 Euro, whereas a 5070 is about 570 Euro, but also weaker than a 9070XT. A 5070Ti, which is about the same performance as the AMD (more or less) starts at 900 Euro. At this point AMD is a nobrainer.
I'm even thinking about downgradin my main monitor to UWQHD or WQHD from UWQXGA (3840x1600) because I can't and not willing to pay a fortune for graphics cards.
My guess is people believe DLSS will continue to be developed and offer better performance in the future. I can't think of a good reason people would pay so much more for similar performance as the AMD offering. Maybe Nvidia GPUs are more common in AI server farms and it's driving the cost of their cards up. Not sure.
It's just within consideration because it's on the market and I simply want to compare all viable products. Intel GPUs aren't in this list because I'm using Linux and their drivers aren't any good.
I'm using nvidia since 780 Ti and got a new one every 2 - 3 gens. I also dislike the decisions nvidia took and its impact on the global hardware market. I hate their AI strategy and hope this bubble finally bursts.
Running a 3840x1600 monitor and 2x 1920x1080 monitors in vertical mode, also in Linux. But I'm only using the big one for gaming.
The initial setup was a bit of a hassle but for the most part because the nvidia drivers were pretty bad.
A lot of people buy Nshitia (me included) because of Cuda and DLSS which are years ahead of AMD and decades ahead of Intel. Even with worse driver support on Linux, Nvidia still outpreforms AMD cards of the same class on every front, except net power consumtion.
From what I've seen they are leading in power consumption as well. Yes, 5090 is more power hungry than any AMD consumer GPU right now but it also comes with adequately higher performance.
With respect to per-watt metrics, 4090 was such an efficiency leap that 5090 struggles to match it. AFAIK 4090 still holds the efficiency crown in discrete graphics. (Maybe not for other acceleration tasks. Also driver/FW updates sometimes shift benchmark standings over time.)
It's fine if you buy it because you know what you need and know that you're paying more because of X. I'm talking about people that buy it just because it's NVIDIA with no thought for anything else.
Probably pretty much exclusively for machine learning applications, though I wonder if it's even still a good choice to use consumer gaming GPUs at these prices. But IDK what a developer would buy instead for home use, I can't quite imagine that something like a Raspberry Pi AI hat has a better performance/cost ratio.
For local inference these days it’s basically all AS since it’s the only consumer hardware that can replace $20-80k server builds, but if you mean ML/AI developers specifically, they still prefer discrete GPUs for local training (on personal budgets: either recent consumer gaming/professional cards or older AI cards).
My bad, Apple Silicon, the private local ML “meta” as of ~Nov’25 (and last I saw) was RDMA clusters of M3 Ultras that could scale up to 3.6TB VRAM (a lot)
Seems to be roughly on par with the RX 6800, according to user benchmarks.com, so everything after should be an improvement. The -XT models are generally stronger. Also, in an attempt to adapt their naming schema to nvidia's amd's 7900 series was succeeded by the 90xx series.
Generally speaking nvidia seems to have the edge in ray-tracing, while amd offers faster rasterisation.
I still have an 3070TI, a 3080 and 2 1080Ti's lying around. I was thinking about selling some of them, but with the current situation i will keep all of them, simply to have spares for years to come.
23 Comments
normanwall@lemmy.world · 43 pts · 20d
Lol This would put the cheapest 5090 near me at $6500AUD, ~$4522USD
Fuck you NVIDIA, happily bought a 9070XT
Senseless@feddit.org · 5 pts · 20d
Just looked for a potential successor for my 3080 and compared 5070, 5070 Ti and 9070XT because they have roughly the same performance. A 9070XT is about 650 Euro, whereas a 5070 is about 570 Euro, but also weaker than a 9070XT. A 5070Ti, which is about the same performance as the AMD (more or less) starts at 900 Euro. At this point AMD is a nobrainer.
I'm even thinking about downgradin my main monitor to UWQHD or WQHD from UWQXGA (3840x1600) because I can't and not willing to pay a fortune for graphics cards.
Mac@mander.xyz · 1 pts · 19d
Why would supporting Nvidia even be in consideration?
Fluffy_Ruffs@lemmy.world · 2 pts · 19d
My guess is people believe DLSS will continue to be developed and offer better performance in the future. I can't think of a good reason people would pay so much more for similar performance as the AMD offering. Maybe Nvidia GPUs are more common in AI server farms and it's driving the cost of their cards up. Not sure.
Senseless@feddit.org · 1 pts · 19d
It's just within consideration because it's on the market and I simply want to compare all viable products. Intel GPUs aren't in this list because I'm using Linux and their drivers aren't any good. I'm using nvidia since 780 Ti and got a new one every 2 - 3 gens. I also dislike the decisions nvidia took and its impact on the global hardware market. I hate their AI strategy and hope this bubble finally bursts.
AnUnusualRelic@lemmy.world · 1 pts · 19d
For what it's worth, I'm running a 5000x1200 (roughly) monitor with a 9070xt with no issues (in Linux).
Senseless@feddit.org · 1 pts · 19d
Running a 3840x1600 monitor and 2x 1920x1080 monitors in vertical mode, also in Linux. But I'm only using the big one for gaming. The initial setup was a bit of a hassle but for the most part because the nvidia drivers were pretty bad.
Chais@sh.itjust.works · 22 pts · 20d
Why would anyone buy nvidia at this point?
WereCat@lemmy.world · 16 pts · 20d
Because a lot of people just buy NVIDIA because it’s NVIDIA still
LouNeko@lemmy.world · 10 pts · 20d
A lot of people buy Nshitia (me included) because of Cuda and DLSS which are years ahead of AMD and decades ahead of Intel. Even with worse driver support on Linux, Nvidia still outpreforms AMD cards of the same class on every front, except net power consumtion.
3bygone3@lemmy.world · 2 pts · 20d
From what I've seen they are leading in power consumption as well. Yes, 5090 is more power hungry than any AMD consumer GPU right now but it also comes with adequately higher performance.
Septimaeus@infosec.pub · 1 pts · 19d
With respect to per-watt metrics, 4090 was such an efficiency leap that 5090 struggles to match it. AFAIK 4090 still holds the efficiency crown in discrete graphics. (Maybe not for other acceleration tasks. Also driver/FW updates sometimes shift benchmark standings over time.)
WereCat@lemmy.world · 1 pts · 17d
It's fine if you buy it because you know what you need and know that you're paying more because of X. I'm talking about people that buy it just because it's NVIDIA with no thought for anything else.
Mac@mander.xyz · 1 pts · 19d
Hope it was worth it
And thanks, btw
rumschlumpel@feddit.org · 4 pts · 20d
Probably pretty much exclusively for machine learning applications, though I wonder if it's even still a good choice to use consumer gaming GPUs at these prices. But IDK what a developer would buy instead for home use, I can't quite imagine that something like a Raspberry Pi AI hat has a better performance/cost ratio.
Septimaeus@infosec.pub · 2 pts · 19d
For local inference these days it’s basically all AS since it’s the only consumer hardware that can replace $20-80k server builds, but if you mean ML/AI developers specifically, they still prefer discrete GPUs for local training (on personal budgets: either recent consumer gaming/professional cards or older AI cards).
rumschlumpel@feddit.org · 2 pts · 19d
What does "AS" abbreviate here?
Septimaeus@infosec.pub · 2 pts · 19d
My bad, Apple Silicon, the private local ML “meta” as of ~Nov’25 (and last I saw) was RDMA clusters of M3 Ultras that could scale up to 3.6TB VRAM (a lot)
CybranM@feddit.nu · 3 pts · 20d
If you have deep pockets and want the best performance Nvidia is the only option
stoy@lemmy.zip · 2 pts · 20d
I have simply kept buying nVidia as I understand what the model numbers mean, I have no idea of what AMDs GPUs model numbers mean...
I have a 3070 at home right now, what would be an upgrade from AMD?
Chais@sh.itjust.works · 6 pts · 20d
Seems to be roughly on par with the RX 6800, according to user benchmarks.com, so everything after should be an improvement. The -XT models are generally stronger. Also, in an attempt to adapt their naming schema to nvidia's amd's 7900 series was succeeded by the 90xx series.
Generally speaking nvidia seems to have the edge in ray-tracing, while amd offers faster rasterisation.
Wildmimic@anarchist.nexus · 5 pts · 20d
I still have an 3070TI, a 3080 and 2 1080Ti's lying around. I was thinking about selling some of them, but with the current situation i will keep all of them, simply to have spares for years to come.
Soulphite@reddthat.com · 1 pts · 20d
"First they came for the dot-com's, and I did not speak out - because I was not a stock trader."
"Then they came for the crypto-mining, and I did not speak out - because I was not a miner."
"Then they came for the Data Centers, and I did not speak out - because I was not a AI tech Bro."
"Then they came for me - and there was no one left to speak for me."