this post was submitted on 31 Jan 2025
71 points (83.8% liked)
Technology
61315 readers
2691 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related content.
- Be excellent to each other!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, to ask if your bot can be added please contact us.
- Check for duplicates before posting, duplicates may be removed
- Accounts 7 days and younger will have their posts automatically removed.
Approved Bots
founded 2 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
Nvidia doesn't give a shit about gamers anymore. The incremental improvements are a side effect. This is why they're so focused on software enhancements instead like DLSS now. It gives them the marketing numbers without having to do the hardware improvements for gaming.
Their bread and butter now is AI, and large scale machine learning. Where businesses are buying thousands of cards at a time. It's also why they're so stingy with VRAM on their cards, large amounts on VRAM are not as necessary for most workloads outside gaming now, and it saves them millions of dollars every generation.
You're right, however I'd say that Nvidia has always been stingy with VRAM. The 3060 had 6GB while the RX 480 had 8GB, for example, the 970 had 3.5GB VRAM and the R9 390 had 8GB, and there are similar examples going back a long way.
It has got pretty bad recently. Worse than normal. AI is also very VRAM intensive (even moreso than gaming), so I imagine they've been diverting those chips to their AI/enterprise cards.
Well, Nvidia seemingly forgot to price gouge on RAM for the 3060 and they had a 12 GB standard version for a while. That should have been the low range standard, with 24 for mid and 32 for high, but they've adjusted.
They only did that because they were forced by AMD's VRAM choices and unexpectedly great RDNA2 architecture.
Because of the memory bus that the 3060 had, it essentially had to have either 6GB of VRAM or 12GB, and it'd have looked stupid next to AMD with only 6GB, so they changed it to 12GB fairly late on in development.
It led to the bizarre situation of the 3060 Ti (based on the 3070 die) having less VRAM at 8GB.
So yeah, less that they didn't want to price gouge, more that AMD was giving 12GB for similarly priced cards that were also much faster, and Nvidia knew that 6GB would look like a joke in comparison.
Thanks, I didn't know that!
I'm just quoting this for emphasis.