52
all 28 comments
sorted by: hot top controversial new old
[-] chandz05@lemmy.world 11 points 1 year ago

Guess I'm keeping my 1070 for another couple years

[-] TowerofPimps@lemmy.world 3 points 1 year ago
[-] MossBear@lemmy.world 11 points 1 year ago

Then I will play retro games until the universe falls to darkness.

[-] Sharpiemarker@feddit.de 2 points 1 year ago

Any day now

[-] RanchOnPancakes@lemmy.world 8 points 1 year ago
[-] JonVonBasslake@lemmy.world 2 points 1 year ago

Not sure if you're asking what they mean by "again", but the prices were just starting to come down after the cryptobros began hording them.

[-] dudewitbow@lemmy.ml 1 points 1 year ago

Nvidia is shifting alottment to AI, they know as long as they hold mind share, people will buy Nvidia over AMD and Intel. AMD adjusts their pricing to Nvidia, but is more prone on getting price drops after launch (e.g how popular the 7900xt was when it price dropped to 700)

[-] UsernameIsTooLon@lemmy.world 7 points 1 year ago

What's the point of really expensive hardware and hard to run games if only a small percentage of people can play them? It's not just that PC parts are getting better, but devs are getting lazier regarding optimizations because they have more headroom than ever. We keep running into new problems like running out of VRAM or shader compilation stuttering. Long live the source engine that has spent more development time on optimizing than actually producing games.

[-] armchair_progamer@programming.dev 4 points 1 year ago* (last edited 1 year ago)

But aren’t the GPUs used by AI different than the GPUs used by gamers? 8GB of RAM isn’t enough to run even the smaller LLMs, you need specialized GPUs with 80+GB like A100s and H100s.

The top-tier consumer models like the 3090 and 4090 have 32GB, with them you can train and run smaller LLMs locally. But there still isn’t much demand to do that because you can rent GPUs on the cloud for cheap; enough that the point where renting exceeds the cost of buying is very far off. For consumers it’s still too expensive to fine-tune your own model, and startups and small businesses have enough money to rent the more expensive, specialized GPUs.

Right now GPU prices aren’t extremely low, but you can actually but them from retailers at market price. That wasn’t the case when crypto-mining was popular

[-] Skua@kbin.social 2 points 1 year ago

Not all applications of it are as demanding as LLMs though. The company referenced in the article as buying likely upwards of twenty million dollars worth of graphics cards seems to work mostly on driver assistance (presumably with the goal of full self-driving at some point, but I haven't looked at them thoroughly). It seems quite possible that however their product works is within the capacity of a GPU that would otherwise be popular for gaming, and it being installed in cars means that it has to have the capabilities in the physical device that's in the car

[-] acedelgado@kbin.social 2 points 1 year ago

They're not that different, really. CUDA processing cores are the most used in AI training, and those are the main processors used in both Nvidia's consumer desktop cards and machine learning enterprise cards. As "AI" is on the rise, more and more of the supply of CUDA processors and VRAM chips will be diverted to enterprise solutions that will fetch a higher price from deals with corporations. Meaning there will be less materials available for the consumer-level GPU supply, which will drive prices up for normal consumers. NVIDIA has been banking on this for a long time; that's why they don't care about overpricing the consumer market and have been trying to push people towards cloud-based GeForce Now subscription models where you don't even own the hardware and just basically rent the processing power to play games.

Also just to be anal, the 3090 and 4090 have 24Gb of vram, not 32Gb. And unlike gaming nowadays you can distribute the workload to multiple GPU's in one system, or over a network of machines.

[-] li10@feddit.uk 3 points 1 year ago

Remember when PC gaming used to be hands down better than console?

Anyone who still believes that is stuck in the past.

PC certainly has its benefits and is my platform of choice, but if somebody was getting into gaming or just casually interested then a PS5 is a much better choice these days.

[-] Delusional@lemmy.world 8 points 1 year ago* (last edited 1 year ago)

Yeah when a single component can cost up to 3x the price of a console then it's no longer worth it. I'd like to upgrade my GPU to get stable 80+fps on all games max graphics but when that would cost me $1000-$1500, I'd rather just play on medium graphics.

[-] alessandro@lemmy.ca 2 points 1 year ago

a PS5 is a much better choice these days.

if you upgrade your PC to "PS5 level of graphic detail" your whole library of games (starting from the original Doom/Quake in ray tracing, upscaled texture pack etc.) get upgraded.

If you buy a PS5... then only the games you buy specifically for the PS5 will run at PS5 graphics.

To this add that, on PC, you can buy second hand GPU that matches PS5 fidelity (both Nvidia or AMD)

[-] HolyDuckTurtle@kbin.social 1 points 1 year ago

The point is that even a second hand high-end card will cost as much as a PS5. For a single component.

PC gaming has grown a lot, but its peak was the PS3/PS4 era when it could be noticably superior to average people. I'd say it's going back to being a platform for entheusiasts who care about those things you mentioned, which most people don't. At least, not enough to spend twice as much or more for a console equivilent.

this post was submitted on 04 Aug 2023
52 points (96.4% liked)

PCGaming

6487 readers
60 users here now

Rule 0: Be civil

Rule #1: No spam, porn, or facilitating piracy

Rule #2: No advertisements

Rule #3: No memes, PCMR language, or low-effort posts/comments

Rule #4: No tech support or game help questions

Rule #5: No questions about building/buying computers, hardware, peripherals, furniture, etc.

Rule #6: No game suggestions, friend requests, surveys, or begging.

Rule #7: No Let's Plays, streams, highlight reels/montages, random videos or shorts

Rule #8: No off-topic posts/comments

Rule #9: Use the original source, no editorialized titles, no duplicates

founded 1 year ago
MODERATORS