Did you forget that on the N64, load times were near instantaneous?

> Apple Vision APIs

Can we even OCR pdfs using Apple Vision APIs

I haven't played with more advanced feature or RAG, but at least AnythingLLM seems to support both local and remote vector databases (you can choose: lancedb, chroma, pinecone, zilliz cloud), embedding provider again both local (ollama, lmstudio) and remote and adjust text splitting and chuncking. Have some Agent skills like Web Search, scraping websites, sql connector. Seems probably the most advanced and I like you can still use ollama or lmstudio as your main LLM models repository.

OpenWebUI a little bit more harder to setup (need docker), but interface very similar to OpenAI

It’s puzzling. Just the other day, it was reported that their power surplus was so significant that rates were even negative. So, why are end-users still paying high prices? Is it because of the distribution costs?

That's tracing garbage collection. Reference counting is another type of garbage collection. https://en.wikipedia.org/wiki/Garbage_collection_(computer_s...

From a quick search suggesting somewhat relevant results:

- Using AMD FSR 2 to upscale a 1280x720 render target to 1080p, and all graphic presets set to low they benchmarked ~42 FPS on M1 Pro and 47 FPS on M1 Max.

Around the same ish ballpark as with a GTX 700 series dedicated GPU, released 10 years ago. And around twice the performance of other integrated graphics of other modern CPUs.

Good point! Have you had any experience with CrossOver? Does it tend to work better than Whisky?

Part of the nuance here is that although both can use GPT underneath, there's still a lot of config that needs to happen beforehand (e.g. different "sync" modes that impact not just visual fidelity, but whether the game will even launch at all). Does CrossOver take care of all that for you (with predefined pre-game profiles, perhaps) or do you still have to manually configure every title?

Much props to the maintainers for making this available at all, but I gotta say that as a user (gamer), the experience is a pain. Some games work, sort of, after a lot of configuration (often with outdated instructions, https://github.com/Whisky-App/Whisky/wiki/Game-Support. Many games don't work at all, or only work on some platforms (e.g. Battle.net but not Steam for Diablo 4). Dozens of issues remain unaddressed: https://github.com/Whisky-App/Whisky/issues

I wish Apple would hire a team to properly maintain this, the way Valve has done with Proton and Steam Deck Verified.

In the meantime, I find that GeForce Now (https://www.nvidia.com/en-us/geforce-now/) provides a much more seamless, one-click-and-play experience. It works every time and doesn't break on every update, etc. Its primary downside (aside from a very reasonable monthly subscription fee) is that it has a relatively small game library, with only a subset of Steam/Epic/Xbox title available to play. But of the games it supports, it provides an overwhelmingly better experience than Whisky, GPT, or even native Apple Silicon gaming. The 4080 is dramatically more powerful than even the highest end Mac, and running it in the cloud means no local noise or heat, and much much better battery life. For anyone interested in actually gaming on a Mac without headaches, with good graphics and a good framerate, I'd strongly recommend GFN and just ignoring the ports or emulation layers. It's just a much better experience.

Tim Cook?