It’s an NVIDIA specific term that is the abbreviation for GPU System Processor. It’s a RISC-V core that does all sorts of management tasks on a modern Nvidia card, mostly related to task setup, resource allocation, context switching, adjusting clock speeds, etc.

I bought a used gen1 Thinkpad X1 Nano. It is super light (<1kg), works flawless out of the box with Linux, and while I think it does have a fan I’ve never noticed it.

Based on the neofetch it’s a Samsung Fold Z 4

I have one of these, but only use it for SteamVR. Does this mean I can’t update either?

AFAIK, the drivers come from Windows.

Edit:

From the article:

Existing Windows Mixed Reality devices will continue to work with Steam through November 2026, if users remain on their current released version of Windows 11 (version 23H2) and do not upgrade to this year’s annual feature update for Windows 11 (version 24H2). This deprecation does not impact HoloLens.

Well fuck. This headset is the only reason I keep a Windows PC around at all.

Works for me in Belgium. Weird.

To quote the author himself:

Great, do whatever you want. Just shut the fuck up about it, nobody cares.

But then he proceeds to do the exact opposite and posts a vitriolic rant about how everyone who doesn’t use what they use is, in their words, and idiot.

This is Indian English, same as “I have a doubt”. It’s not something you commonly hear outside India.

I’m not trying to be defeatist. I’m just advocating for unions to stop focusing on the local aspect. Transnational cooperation is what is needed in unions today.

Something which larger unions, such as the steel and car industry unions in Europe have been trying (and mostly failing) to do for almost two decades.

Companies, by and large, have used the globalized economy to sidestep local action for almost 30 years now.

Ignoring this is simply a recepy for repeating the mistakes of the past. Especially in software, where there is no physical production equipment at all, and in games, where talent and labour is plentiful.

Unless you have an organization that reaches as far as the companies you’re trying to bring to the table, you will simply be outmanœuvred.

You also overestimate the level of union participation if you think they would need to lay off everyone to break a union strike.

Feel free to tell me I’m wrong, but I’ve been through the proces twice, both involving union action, both in the software industry. Once as part of the workers delegation to the negotiating table. Dismiss a company’s ruthlessness or resourcefulness at your own peril.

Local unions can only hope to hold off the axe until current projects where the required know how can not be rebuilt or transfered in time are done.

One issue is that unions have failed to globalize while industries have. CDPR could simply chose to bypass the union by opening a dev studio in a country with no or less union presence.

Given the recent wave of layoffs in the game industry, they’ll have no shortage finding capable people.

Seeing the runaway succes of others like Nvidia, Apple, Mediatek, do you think any meaningful new entries are going to deviate from their playbook?

Being a good citizen with regards to transparency in firmware and Linux support is not a proven differentiator for these vendors and shown time and time again not to be a requirement for success.

I don’t understand what you mean. Why does ARM hardware become obsolete after a few years? Lacking ongoing software support and no mainline Linux?

What does that have to do with the instruction set license? If you think RISC-V implementors who actually make the damn chips won’t ship locked hardware that only run signed and encrypted binary blobs, you are in for a disappointing ride.

Major adopters, like WD and Nvidia didn’t pick RISC-V over arm for our freedoms. They were testing the waters to see if they could stop paying the ARM tax. All the other stuff will stay the same.

You give him far too much credit. Never attribute to malice what can adequately be explained by stupidity.

And does this still work with modern cpufreq schedulers like amd-pstate and the Intel equivalent? IIRC I couldn’t simply set frequencies or select the userspace scheduler on 10th gen Intel and frequencies don’t seem to be honoured by AMD pstate drivers on Zen4.

Well, you lose a lot of power efficiency, this would be massively detrimental to many peoples experiences if you do this on anything battery powered like a laptop.

This used to be me but mostly because I would experiment a little too much, never without reason.

Except a few Arch updates over a decade ago when they changed the default from hal to udev, or a Gentoo setup with WAY too specific USE flags, I don’t think I can remember any failure like this ever. I’ve honestly had more issues with Windows nuking itself on a major update.

Mostly using Debian and Fedora these days, and it’s been smooth sailing for quite some time.

Given how little spotify gives to artists, I can’t imagine this being a cost effective way to launder your money at all.

I guess these guys haven’t tried Debian.

While C is certainly better for some problems in my experience, it too is very hard to use in large projects with a mix of developers, and it is unsuitable for most higher level applications in most companies.

I think C has its place in the world still, which is mostly confined low level embedded, kernel space and malware. I do believe that the market segment that used to rely on C++ is today better served by either Go or Rust, depending on the project.

That said, while I LOVE working with Rust, it suffers from many of the same issues I mentioned for C++ in my comment above when working in a mixed skillset team.

Everything is fine within the scope of a college course or project.

Where C++ breaks down is large, complicated projects where you colaborate with other developers over multiple years.

I worked in C++ for almost a decade, and while there were a few good projects I encountered, most suffered from one or more of the following problems:

• C++ has so many parts, everyone picks a subset they think is “good”, but noone seems to fully agree on what that subset is.
• A side effect of the many possibilities C++ offers to compose or abstract your project is that it allows for developers to be “clever”. However, this often results in code that is hard to maintain or understand, especially for other developers.
• Good C++ is very hard. Not everyone is a C++ veteran that read dozens of books or has a robust body of knowledge on all its quirks and pitfalls, and those people are also often assigned to your project and contribute to it. I was certainly never an expert, despite a lot of time and effort spent learning and using C++.

Sorry, I should have included an example in my comment to clarify, but I was in a rush.

HMAC is a widely used technique relies on hashing of a shared secret for verifying authenticity and integrity of a message, for example.