I didn’t think of that, but it’s true! 😄

Worth pointing out that MIPS is a mere shadow of it’s former self (some of my first routers were based on MIPS). They are currently a relatively small RISC-V design firm.

I do wonder how much of IBM’s POWER business is based on net new installations.

If you are running a a critical system on a POWER system (e.g. banking core), it makes sense to pay a high price and avoid the risks associated with complete rewrite.

But does IBM get any new client for their POWER systems?

With a quote post, the post you are quoting is fully (text, avatar image) nested in your post.

It was commonly used on Twitter (I deleted my accounts some time ago).

This would be a really neat feature and would allow for the best of both worlds so to speak.

Yes, I didn’t mean cryptography of course. :)

There are some very useful use cases. One that I have a decent amount of experience in is upscaling SD (or even VHS) content. Depending on the quality of the source material, you can get very good results.

I’ve also found LLMs helpful as a compliment to web searching for my work (I deal with lots of public datasets from international organisations); LLM queries have helped me find sources that I missed via directed search.

It can also be helpful as guided learning/reference system for Linux CLI (I tend to forget the parts I rarely use) or even software application more broadly (used it to help learn about GIS applications that I needed to use to access historical weather data for a work project).

This is pretty cool, I would consider buying a pre-made kit for such a device, especially if the resolution was higher than 768.

No idea, I do use Mastodon relatively regularly, but I am not on top of development discussions.

This was news for me so I decided to share. :)

I am assuming the fix for the second vulnerability will be backported to the older version in bookworn; 13p1 if I remember correctly.

I’ve seen theme backport security to older releases of much less important software.

Pretty good article; a nice overview of recent developments.

I also agree with their conclusion; avoid American commercial entities whenever possible. The culture of corruption and criminality is simply too deeply rooted among the executive class and it is highly unlikely that there will be any true anti-corruption reform in the next ~20 years (I hope I am wrong).

No idea who the Maguire fellow is, but since he works for a large American VC firm, the piece about “coming from a culture of lying” is clearly projection.

I wouldn’t go that far, both LLMs and other ML technologies clearly have novel use cases (unlike crypto where the only use cases are financial speculation and crime).

The issue is that AI promoters/conmen are knowingly misrepresenting the capabilities of their services to benefit from pump and dumps (case in point Microsoft and OpenAI agreeing to define AGI as any service that gets $100 B in annual revenue).

This wouldn’t apply to Debian derived distros (e.g. DietPi). I am assuming the fix will backported to the version of sudo shipping with bookworm.

With that in mind, the DoE hopes microreactors could be used to power small, remote sites, with an eVinci described as a possible power source for a remote datacenter. The Radiant Kaleidos, meanwhile, is described as an alternative to a diesel generator, being a similar tractor-trailer-mounted size while being able to operate for five years without needing refueling.

Alternative to diesel generator does not inspire confidence.

I support nuclear power, but not litering the landscape with random small scale reactors that may or may not be well contained.

I wish this was true, but it seems to be more tied to polticial matters affecting supply chain.

the eVinci is designed to only produce 5 megawatts, and the Kaleidos is limited to just 1.2 megawatts of electrical power output.

Why would you even develop nuclear solutions with such a low output? There is no way people are going to be OK with casually moving these trailers (e.g. to a datacentre). If you’re using nuclear, you might as well go big and develop gigawatts of power.

I could see book-style foldables becoming more popular if the pride goes down, I thought they were useful in showroom testing.

But you are right, prices are still very high.

This is pretty cool (in a very nerdy way of course :) ).

Smartphone hardware is generally noticeably weaker than laptop (let alone desktop) hardware.

There is not much you can do about this.

486 were generally better up until Pentium 2 came out.

No way. Late stage Pentiums were markedly better than 486.

My parents bought a Pentium 1 133 mhz with Windows 95 in late 1996 and it worked great (for the time)

At my parents work they had an 486 with Windows 95 and it was noticeably less responsive. It was not a fun experience. I don’t remember the exact model number of the 486 (I wasn’t even 10 years old back then), but I do remember Pentiums being generally much more performant than computers with 486. From memory, 486 computers generally ran Windows 3.11 and were not upgraded to Windows 95 (keep in mind that everything was pirated, licensed Windows copies among business users was more of mid-2000s thing where I lived).