65 watts will be fine without the dGPU module. If anything, the CPU will be limited to a slightly lower wattage (see here). EDIT: this seems to be a bug.

x86/x64 code is pretty much 100% compatible between AMD and Intel. On the GPU side it’s not that simple but Sony would’ve “just” had to port over their GNM(X) graphics APIs to Intel (Arc, presumably). Just like most PC games work completely fine and in the same way between Nvidia, AMD and Intel GPUs. But they have to do that anyway to some extent even with newer GPU architectures from AMD, because PS4’s GCN isn’t 1:1 compatible to PS5’s RDNA2 on an architectural level, and the PS4’s Jaguar CPU isn’t even close to PS5’s Zen 2.

Other than that, you’re right. Sony wouldn’t switch to Intel unless they got a way better chip and/or way better deal, and I don’t think Intel was ready with a competitive GPU architecture back when the PS5’s specifications were set in stone.

The “Apple TV” is Apple hardware.

How do you not do that? It’s all in your local network, how would it not work offline…?

Replace the 3060 with an equally-priced AMD card and you’ll actually get something decent for your money. Nvidia is horrible at these “lower” price points.

It is, but then again many (most) are hosted on GitHub.

Will reflash my Deck once this hits stable as it’s acting up in various different ways.

Yeah, duplicate flags should just be ignored.

Catppuccin Latte works well too!

To be fair, a big portion of the work that goes into Linux (at least the kernel) is done by paid developers working for big corporations.

I think it’s mostly supply/demand.

Most people are satisfied with how games are acquired commercially. Steam’s DRM system is usually received well. There are outliers using different launchers (sometimes on top of Steam) or games using Denuvo, but most customers are satisfied with how Steam handles it, and it also adds valuable features like cloud saves (so for example when you have a desktop PC and a Steam Deck resuming where you left off is pretty seamless) and Valve didn’t have any major fuckups yet (not that I remember anyway). It works, it’s convenient and most people can afford it.

Similar thing with music: streaming services work well for the most part and have almost all the music most people would want. They’re pretty affordable and convenient.

With movies and TV shows most people were satisfied when Netflix got rolling as it was pretty much the only streaming service you “needed”. Nowadays more and more services emerge with their own exclusive content and pricing is increased on a regular basis, sometimes multiple times per year. That’s why (from my perspective at least) piracy increases in that sector. It’s no longer affordable and no longer convenient.

As for software, I think most people exclusively use free-to-use software anyway. Software from the Adobe suite still gets pirated a lot, I know no one who paid for Adobe software for personal use.

Considering Intel is behind TSMC as well, China might be quite close to Intel then.

More than enough for Apple to bend to pretty much everything the Chinese government is asking for.

The difference between H.265 and AV1 at the same bitrate (assuming both files were encoded with a good encoder) usually isn’t huge.

AV1 is great, but the “hype” surrounding it is mostly comparing it to lowish-bitrate H.264 (live) streams.

Article from JUNE 14, 2022

Article from JUNE 14, 2022

Even if all of that worked, why would I want it?

iOS used to be able to handle background tasks via very specific APIs, that started with iOS 4 and I believe this started to be reworked with iOS 7 and it behaves similar to Android in that background apps are suspended by default. According to an old video by Android Authority, iOS seems to be able to compress suspended apps down to a smaller memory footprint than Android. Both OS allow background services to run, but to my understanding iOS keeps way more control over that compared to Android (although vendor-specific battery saving features probably attempt to do something similar on Android). So in that way, it’s still more specific/selective on iOS compared to Android. Prompt (iOS SSH app) uses the location service in the background to prevent iOS eventually killing active connections for example. Still, iOS seems to handle app suspension more efficient than Android (and yes, Android actually suspends background apps as well).

I’m with you that they could’ve likely bumped all soon-to-be-released iPhone 16 models to 16 GB, but rumors only have them at 8 GB. Makes “sense”, as even the iPad Pro and MacBook Air still only come with 8 GB in their lowest configurations.

But I don’t buy that them releasing the iPhone 15 with only 6 GB of RAM was a malicious attempt at limiting AI features. Seeing how unfinished their AI stuff is even in their latest beta releases, they were/are playing catchup. It was bad foresight and there are often talks about how internal teams at Apple are very secretive about projects in development, I wouldn’t be surprised if the team developing the iPhone 15 knew pretty much nothing about the software plans with Apple Intelligence. It’s still a very valid point of criticism though obviously, seeing as you could still buy an iPhone 15 to this day (it’s still the “latest and greatest” non-Pro iPhone before the iPhone 16 releases in a few weeks) and you won’t get the by far biggest feature of a software update releasing just weeks/months after your purchase. This is a huge step backwards in terms of software support, as iPhones normally get pretty much all major new software features for at least 3 years, and still most features of even newer OS releases (recent devices have seen support for major updates for 6+ years, the iPhone XS will get its 7th major iOS release with iOS 18).

I’m not saying “cut that poor multi-trillion dollar company a break”, I’m just saying that not supporting the iPhone 15 for Apple Intelligence probably isn’t a result of malicious acting, but rather bad foresight and poor internal communication. Limiting the soon-to-be-released iPhone 16 models to 8 GB on the other hand seems very greedy, especially with them trying to run as many of their AI models on-device.

And what would that help with?

See if you can find a (used) GPD Win (Max).

Otherwise maybe a used (~2016) 12" MacBook.

Other than that maybe some DIY solutions.

It’ll be cheaper if the device can be a little bigger, plenty of dirt cheap used 13" laptops out there.