Sorry if I’m not the first to bring this up. It seems like a simple enough solution.
A nice place to discuss rumors, happenings, innovations, and challenges in the technology sphere. We also welcome discussions on the intersections of technology and society. If it’s technological news or discussion of technology, it probably belongs here.
Remember the overriding ethos on Beehaw: Be(e) Nice. Each user you encounter here is a person, and should be treated with kindness (even if they’re wrong, or use a Linux distro you don’t like). Personal attacks will not be tolerated.
Subcommunities on Beehaw:
This community’s icon was made by Aaron Schneider, under the CC-BY-NC-SA 4.0 license.
Yeah, but cuda tho
In this particular case, it’s a bit more complicated.
I suspect the majority of 30x0 & 40x0 card sales continue to be for non-gaming or hybrid uses. I suspect that if pure gamers stopped buying them today for months, it wouldn’t make much of a difference to their bottom line.
Until there’s reasonable competition for training AI models at reasonable prices, people are going to continue buying their cards because it’s the most cost-effective thing – even at the outrageous prices.
China is still buying gimped cards because what other choice do they have
Exactly. The only other choices here are to buy used with a risk or wait longer to upgrade.
Well when AMD finally catches up with nVidia and offers a high-end GPU with FG and decent Ray Tracing, I’ll gladly switch. I’d love nothing more than to have an all-AMD PC.
I was originally eyeballing to get 3080 pre-covid then everything goes to the moon with crypto, can’t even buy 2080 under 1k. So I stuck with my 1080, some shit happened to my MB so I had to upgrade and that’s where I decide to switch to all AMD, took me another couple months before I finally snatch a 6800XT at amd direct, it was not cheap($812 CAD after tax) but like almost 500 cheaper than what you can find on any other vendor tries to gouge you with their bundles.
Quite happy with that and their software. BUT, there are some weird crash related the screen sleep after certain version of driver so I turn off screen manually. A newer driver seems to fix that about a month or 2 ago. I will wait a couple update to revert my screen sleep policy again. Most modern game runs quite well.
Honest the newer 7900 ones are priced too high so I might either skip entire gen or wait until 8900 is out then get a good 7900 XT for cheaper price. Basically I have no intention to buy any thing priced higher than 800 pre-tax.
With cheap ram and everything else, there is really no reason for GPU to still keep at that price point. Remember 3080 was priced $699? Yep, that’s where I get my budget number of <800CAD.
I’m looking at buying a gaming laptop, I have yet to find anything worth buying with an AMD card.
Framework 16
That’s a decent option, but I could get a more powerful laptop with an Nvidia GPU for cheaper from System76.
deleted by creator
The topic at hand is about Nvidia price gouging, so the AMD option being more expensive than the Nvidia option is certainly relevant.
Sure, but this has the AMD you requested, plus will last you much longer than the System76, since you can repair and upgrade it.
Correct.
Uhh they did. And the card prices are still high, so what now, genius?
Not getting enough sales, gotta jack up the price so they make the same amount of money. Seems legit.
You know that, I know that. Most people browsing here know that. Anyone you speak to who buys GPUs would probably also agree. Still not gonna change.
What other company besides AMD makes GPUs, and what other company makes GPUs that are supported by machine learning programs?
No joke, probably intel. The cards won’t hold a candle to a 4090 but they’re actually pretty decent for both gaming and ML tasks. AMD definitely needs to speed up the timeline on their new ML api tho.
Problem with Intel cards is that they’re a relatively recent release, and not very popular yet. It’s going to be a while before games optimize for them.
For example, the ARC cards aren’t supported for Starfield. Like they might run but not as well as they could if Starfield had optimized for them too. But the card’s only been out a year.
The more people use Arc the quicker it becomes mainstream and optimised for but arc is still considered “beta” and slow in peoples minds even though there were huge improvements and the old benchmarks don’t hold any value anymore. chicken and Egg problem. :/
Disclaimer: i have an arc 770 16GB because every other sensible upgrade path would have cost 3x-4x more for the same performance uplift (and I’m not buying an 8GB card in 2023+) but now I’m starting to get really angry at people blaming Intel for “not supporting this new game” - all that gpus should support is the graphics API to the letter of the specification, all this day-1 patching and driver hotfixes to make games run decent is bs. Games need to feed the API and GPUs need to process what the API tells it to, nothing more nothing less. It’s a complex issue and i think Nvidia held the monopoly for too long, everything is optimised for Nvidia at the cost of making it worse for everyone else.
Isn’t the entire point of DirectX and OpenGL that it abstracts away the GPU-specific details? You write code once and it works on any graphics card that supports the standard? It sounds like games are moving towards what we had in the old days, where they have specific code per graphics card?
I think the issue started with gpu-architecture tailored technologies like physx or gameworks but im probably wrong. For example I have nothing against physx but it only runs on nvidia cores natively (fast), i have an issue when there’s a monetary incentive or exclusive partnering of nvidia and game studios - so if you want to play the game with all the features, bells and whistles, it was designed with you would need to also buy their overpriced (and current gen: underperforming) gpus just because you’d be missing out on features or performance on any other gpu architecture.
If this trend continues everybody will need a €1k+ gpu from nvidia and a €1k+ gpu from AMD and hot-swap between them depending on what game you wish to play.
I jumped to team red this build.
I have been very happy with my 7900XTX.
4K max settings / FPS on every game I’ve thrown at it.
I don’t play the latest games, so I guess I could hit a wall if I play the recent AAA releases, but many times they simply don’t interest me.
Exactly, Nvidia doesn’t have real competition. In gaming sure, but no one is actually competiting with CUDA.
AMD has ROCm which tries to get close. I’ve been able to get some CUDA applications running on a 6700xt, although they are noticeably slower than running on a comparable NVidia card. Maybe we’ll see more projects adding native ROCm support now that AMD is trying to cater to the enterprise market.
They kinda have that, yes. But it was not supported on windows until this year and is in general not officially supported on consumer graphics cards.
Still hoping it will improve, because AMD ships with more VRAM at the same price point, but ROCm feels kinda half assed when looking at the official support investment by AMD.
I don’t own any nvidia hardware out of principal, but ROCm is no where even close to cuda as far as mindshare goes. At this point I rather just have a cuda->rocm shim I can use, in the same was as directx->vulkan does with proton. Trying to fight for mindshare sucks, so trying to get every dev to support it just feel like a massive uphill battle.
My Intel Arc 750 works quite well at 1080 and is perfectly sufficient for me. If people need hyper refresh rates and resolution and all all the bells well then have fun paying for it. But if you need functional, competent gaming, at US$200 Arc is nice.
AMD supports ML, its just a lot of smaller projects are made with CUDA backends, and dont have developers there to switch from CUDA to OpenCL or similar.
Some of the major ML libraries that used to built around CUDA like Tensorflow has already made non CUDA branches, but thats only because tensorflow is open source, ubiquitous in the scene and litterally has google behind it.
ML for more niche uses basically is in the chicken and egg situation. People wont use other gpus for ML because theres no dev working on non CUDA backends. No ones working on non CUDA backends because the devs end up buying Nvidia, which is basically what Nvidia wants.
There are a bunch of followers but a lack in of leaders to move the direction in a more open compute environment.
Huh, my bad. I was operating off of old information. They’ve actually already released the sdk and apis I was referring to.
Apple. Their own processors have both GPUs and AI accelerators. But for some reason, the industry refuses to use them.
I’m doing my part! (with an all-AMD setup)
Similarly if nvidia wanted me to buy their cards they’d get their drivers sorted out. While plugging in an amd card just works with literally no setup from me, nvidia will get no money from me
Funnily enough I just, like an hour before reading this post bought an AMD card. And I’ve been using NVIDIA since the early 00’s.
For me it’s good linux support. Tired of dealing with their drivers.
Will losing me as a customer make a difference to NVIDIA? Nope. Do I feel good about ditching a company that doesn’t treat me well as a consumer? Absolutely!
Have a 3060ti, was thinking of moving to Linux. Is there no support from Nvidia?
You’ll almost certainly be perfectly fine. AMD cards generally work a lot smoother, and the open source drivers means things can be well supported all the time and it’s great.
On Nvidia, in my experience, it’s occasionally a hassle if you’re using a bleeding edge kernel (which you won’t be if you’re on a “normal” distro), where something changes and breaks the proprietary Nvidia driver… And if Nvidia drops support for your graphics card in their driver you may have issues upgrading to a new kernel because the old driver won’t work on the new kernel. But honestly, I wouldn’t let any of this get in the way of running Linux. You have a new card, you’ll probably upgrade before it’s an issue, and the proprietary driver is something we all get mad about, but it mostly works well and there’s a good chance you won’t really notice any issues.
Nvidia on Linux is better than ever before, even over the past couple of months there were tremendous improvements. As long as you use X11 you will have a pretty good gaming experience across the board, but Nvidia driver updates are often a headache. With AMD, you don’t even have to think about it, unless Davinci Resolve forces you to, but even then it’s a better situation.
Anyway, comments like “you’ll be 100% fine” are not really based on reality, occasionally Nvidia will break things. However if you use the BTRFS filesystem with Timeshift (or even that wretched Snapper) set up, then this is merely a minor inconvenience. (for example before I moved to Arch, Ubuntu pushed an Nvidia update that broke my system, happened in June…)
I ran my 1060 just fine for a few year. Nvidia has an official, but proprietary driver that might not run well on some distro’s. Personally I haven’t had any issues, though it would be better to stick with xorg and not wayland. Wayland support on nvidia I’ve heard isn’t great, but it does work
This. You’re mostly at the mercy of their proprietary drivers. There’s issues, like lagging Wayland support as mentioned. They will generally work though, I don’t want to dissuade you from trying out Linux.
There is an open source driver too, but it doesn’t perform well.
Depends on the distro. Otherwise you’ll have to install the nvidia drivers yourself, and if memory serves it’s not as smooth of a process as on Windows. If you use Pop OS you should be golden, as that Linux distro does all the work for you.
Absolutely indeed! I’ll never buy an Nvidia card because of how anti-customer they are. It started with them locking out PCI passthrough when I was building a gaming Linux machine like ten years ago.
I wonder if moving people towards the idea of just following the companies that don’t treat them with contempt is an angle that will work. I know Steph Sterling’s video on “consumer” vs “customer” helped crystallize that attitude in me.
I’m not familiar with that video but I’m intrigued. I’ll have to check it out.
I don’t know. I don’t have much faith in people to act against companies in a meaningful way. Amazon and Walmart are good examples. I feel like it’s common knowledge at this point that these companies are harmful but still they thrive.
Suddenly your video card is as mundane and trivial a solved problem as your keyboard or mouse.
It just works and you never have to even think about it.
To even consider that a reality as someone who’s used Linux since Ubuntu 8.10… I feel spoiled.
Those were rough days. I started with Dapper Drake but there was no way to actually get my trackpad drivers until 8.04. Kudos for sticking with linux
I was hooked. It was the first time my PC felt as transparent and lie-free as notebook paper.
Like, there’s nothing to hide because nothing is. It’s pure, truthful freedom and that meant more to me than raw usability. I tried to do everything possible on Linux that i was told I couldn’t do, hell, I ran Team Fortress 2 and Half Life in wine way pre-proton.
and it sucked, but it was cool tho!
Don’t even get me started on linux audio support.
I recall exactly once back in the day that Ubuntu actually just played audio through a laptop I installed it on and I damn near lost my mind.
like 30 minutes ago I installed Mint on a laptop and literally everything just worked as if I installed windows from the backup image. (I’m not sure power states are working 100% but it’s close enough and probably would with 3rd party driver)
My wife and I have a joke. It is always the audio.
We have used Linux since about 1998 and more seriously since the early 2000’s. Recently not had audio problems but I am not sure they are done with. Historically every time I think Linux has solved audio, there is always another problem it seems.
I used some Ubuntu derivative for recording shitty music me and my buddy made in a trailer. OSS off of a turtle beach soundcard with a hacked together driver, crammed into a shitty Windows Vista era desktop.
I felt like some sort of junk wizard.
I use arch these days, Garuda mainly. I’ve done the whole song and dance from Arch to Gentoo. I know the system, now I want to relax and let something I suck at, giving myself features be more in the hands of a catering staff of folks and the Garuda boys know how to pamper.
The dragons kinda… yeah, the art’s kinda cringe but damn, this is the definition of fully featured.
I was definitely a junk wizard back in the day, as I’ve grown older and have less time and more money I just want stuff that works. I used to build entire (pretty acceptably decent) home theater systems out of $150 worth of stuff off craigslist and yard sales. When you know how it all works you can cobble together some real goofy shit that works.
It’s about the exact amount of cringe I expect from a non mainstream linux distro. but aye who doesn’t like dragons and eagles? I’ll have to try it out on this old zenbook.
As a console gamer, I don’t have to worry about it. Xbox Series X, PS5, Steam Deck are all AMD based. The Switch is Nvidia, but honestly, I can’t remember the last time I turned the Switch on.
They have enough datacenters buying their products, they don’t need consumers so they can wait more than the average user…
I personally am done with gaming because if I must spend a thousand euros, I’ll spend it in bicicle parts and accessories and not on one single piece of hardware for a decent PC.
Why would datacenters be buying consumer grade cards? Nvidia has the A series cards for enterprise that are basically identical to consumer ones but with features useful for enterprise unlocked.
They don’t but Nvidia diverts some of its production towards gaming chips. If nobody buys graphics cards they can easily keep a stash of those chips to sell down the line and go 100% on datacenter products.
I think you mean their Tesla line of cards? The A (e.g. A100) stands for the generation name (e.g. Ada or Ampere, don’t remember which one got the A), and that same name applies to both the consumer line (GeForce and Quadro) and the data’s centre cards.
The hardware isn’t identical either. I don’t know all the differences, but I know at least that the data centre cards have SXM connectors that greatly increase data throughput.
There are also games that don’t render a square mile of a city in photorealistic quality.
I’m currently part of the problem and this is so fucking true. Games have really stopped pushing the envelope because they either have to be cross platform compatible or they’re not even PC first.
3D mark is the only thing I could find to put a dent in my 3060ti
Graphical fidelity has not materially improved since the days of Crysis 1, 16 years ago. The only two meaningful changes for how difficult games should be to run in that time are that 1440p & 2160p have become more common, and raytracing. But consoles being content to run at dynamic resolutions and 30fps combined with tools developed to make raytracting palatable (DLSS) have made developers complacent to have their games run like absolute garbage even on mid spec hardware that should have no trouble running 1080p/60fps.
Destiny 2 was famously well optimized at launch. I was running an easy 1440p/120fps in pretty much all scenarios maxed out on a 1080 Ti. The more new zones come out, the worse performance seems to be in each, even though I now have a 3090.
I am loving BG3 but the entire city in act 3 can barely run 40fps on a 3090, and it is not an especially gorgeous looking game. The only thing I can really imagine is that maxed out the character models and armor models do look quite nice. But a lot of environment art is extremely low-poly. I should not have to turn on DLSS to get playable framerates in a game like this with a Titan class card.
Nvidia and AMD just keep cranking the power on the cards, they’re now 3+ slot behemoths to deal with all the heat, which also means cranking the price. They also seem to think 30fps is acceptable, which it just… is not. Especially not in first person games.
I think you may have rose tinted glasses on this point, the level of detail in environments and accuracy of shading, especially of dynamic objects, has increased greatly. Material shading has also gotten insanely good compared to what we had then. Just peep the PBR materials on guns in modern FPS games, it’s incredible, Crysis just had normals and specular maps all black or grey guns that are kinda shiny and normal mapped. If you went inside of a small building or whatever there was hardly any shading or shadows to make it look right either.
Crysis is a very clever use of what was available to make it look good, but we can do a hell of a lot better now (without raytracing) At the time shaders were getting really computationally cheap to implement so those still look relatively good, but geometry and framebuffer size just did not keep pace at all, tesselation was the next hotness after that because it was supposed to help fix the limited geometry horsepower contemporary cards had by utilizing their extremely powerful shader cores to do some of the heavy lifting. Just look at the rocks in Crysis compared to the foliage and it’s really obvious this was the case. Bad Company 2 is another good example of good shaders with really crushingly limited geometry though there are clever workarounds there to make it look pretty good still.
I could see the argument that the juice isn’t worth the squeeze to you, but graphics have very noticeably advanced in that time.
W/r/t Baldur’s Gate 3, I don’t think the bottleneck is the GPU. Act 3 is incredibly ambitious in terms of NPC density, and AI is one of those things that’s still very hard to parallelize.
I went with best AMD I could get at the time, 7900xtx sapphire nitro. For gaming, it’s already really good, I can use raytracing, although not on best settings on some games,but in most cases I can just max out the settings.
Main complaint atm is that self hosting AI is much more annoying than it would be on nvidia. I usually get everything to work eventually, but it always needs that extra level of technical fuckery.