Don’t tend to have a terminal emulator of any kind installed on remote boxes. They’re headless.

That exception is my primary use case for tmux, so that explains it.

As a non-user of kitty, why did it make you drop tmux? Don’t they do different jobs?

Those are rookie numbers.

Yes. Yes you can.

Ignore it. Move on.

Your 4 drive raid5 array, right?

Right?!

BLM and ACAB are very American centric movements/sentiments. Hence why I took your comments to be American focused.

Whilst I’ve heard the phrase “thin blue line” before, it’s never been something associated with racist overtones or subjugative ones in my experience. More that the police is a small community protecting the larger population.

As such, someone using it context of not allowing the proliferation something is reasonable, if possibly histrionic in this case.

You’re talking about the US police. A lot of the world have police forces that serve the people.

When the questions you ask chatGPT even offend the browser you’re using!

It was fine when rendering (esp. text) was server side and not client side like it is now. At least LAN (10MB ethernet) was basically transparent. Internet was shit mainly because everyone was on 56k modems.

GTK and Qt do all their rendering client side and transfer bitmaps to the server requiring much more bandwidth.

Looking at the Product Conformance pages, it looks like you need a GCN 1.0 device for Vulkan 1.0 support (current is 1.4).

Thames was a Terrascale 2 architecture (i.e. the one before GCN1.0) and as such can’t support Vulkan or DX12.

The rationale for doing this must only be for fun. You’re talking about a project that will take man years, and the price of a new vulkan capable graphics card is $300-400.

Just learn the basic POSIX commands (there’s probably 20-25) and understand pipes. Then you can do pretty much anything you’re likely to need from the command line. Sure, there will be more modern flashy friendly tools that come along. Some you’ll integrate into what you do. Some won’t have enough staying power to remember.

No notes. No googling or LLM. Just don’t skip to the end.

I can see lots of subtle bugs coming from this.

Develop and everything working great. Deploy on server with different HWCaps and you’re now running different code.

enabling unmodified CUDA applications to run on AMD GPUs at near-native performance, the ZLUDA atop AMD HIP code was made available and open-source following the end of the AMD contract

Trouble is… HIP doesn’t support all of AMDs GPUs. It’s only 7900s in the consumer line-up.

Thanks for coming back and letting others know what your solution was.

Boot from a USB stick with a Live environment on it. See if you get the same issue.

For most people, using Linux is not a buggy experience. So no, people aren’t gaslighting you. Normally, you grab a modern release like the latest Fedora or Ubuntu and you can get a live desktop up in seconds booting from a USB stick.

Esoteric hardware can be a problem if particular driver haven’t been developed yet. That tends to hit laptops harder than desktops, but it’s much less of an issue than it used to be.

People are asking for specifics because they don’t share your experience and so can’t fill in the blanks.

Probably never. Those 2048 cores are all separate threads. GPUs work with a cluster of cores all following a single thread of execution.

MIMD Vs SIMD.

Not OP, but I doubt that’s what he meant. An APU still has a CPU and a GPU on it as separate things.