

The Framework CEO said in a QA that AMD engineers did a simulation of this APU with CAMM and that they’d have to cut the memory bandwidth by half to make it work.
The Framework CEO said in a QA that AMD engineers did a simulation of this APU with CAMM and that they’d have to cut the memory bandwidth by half to make it work.
I guess we should also force everyone to use wheelchairs because they are great for paraplegics.
Even “heavy” linux distros are easy to run for a pc capable of running and streaming modern games. Lightweight distros are mostly relevant for older or underpowered hardware.
Have you checked out https://openrazer.github.io/#devices
The easy path to getting storefronts like EA working is through Lutris. It does all the setup for you through guided wizards. I can’t help much with deciding a distro tho, I’ve been using Fedora for years and that works well enough but is not exactly gaming focussed.
There are benchmarks on the huggingface page. The larger model is close to GPT4o performance. Which makes this worse than deepseek-r1. But it is a smaller model and not a reasoning model (doesn’t use up extra tokens to “think”). So still very impressive and important for open source.
Belgian here. We have not fallen to fascism. I would not call NVA extreme right, just right. That would be Vlaams Belang, who also got a lot of votes to be fair but they’re not part of the government coalition because none of the other parties want to work with those nutcases. There is a coalition of parties. Because we have a healthy number of parties to vote for, no single party ever gets the majority vote. Five parties just signed a coalition agreement. That’s why Belgium has the record for being the longest without a government. It takes a long time for these parties to come to an agreement but it also means more balanced policies. Things change slowly in Belgian government, we don’t get big swings from left to right. But that doesn’t make for good headlines.
Playing devil’s advocate here. Mouse movements and key presses have been commonly used as bot detection method for a decade now. Like that captcha service that is just a checkbox, that’s part of how they guessed that you are not a bot.
Training data for these models used to be text off of the internet and some manually generated Q&A examples to make it behave more like a chat bot (instruction tuning). Because there is still a need for more data they have started adding AI generated text to the dataset. This technique doesn’t add new knowledge but it has shown to reduce hallucinations. Likely because this data is more focussed, truthful and structured than the median text from the existing datasets. They would probably have data from every major chat provider in there, especially the big boys.
Frame gen is interpoplation. It does introduce latency. But nvidia has Reflex to offset some of that
In case you forgot, Lemmy is social media
A privately owned platform cannot serve the public good. There will always be conflicts of interest. A proper public square should be funded by a competent government (but those are rare) or decentralized.
Ah good to know. I was just thinking maybe the OP was a spammer, got cancelled or something
Wait what’s wrong with this blog post? I’m finding it very interesting and well written
Sir I think you have stumbled into the wrong sublemmy