|
Advantech has released the EAI-3100, converting Intel's Arc A370M mobile graphics card into one for the desktop. Get all the graphics processing power of a laptop in your desktop!
|
|
|
|
|
Microsoft is testing native Sudo command support for Windows 11. We'll finally be able to get that sandwich!
For those that don't get the 'joke': xkcd: Sandwich[^]
|
|
|
|
|
Why do I think that once it is broadcasted through the updates, the news about security flaws in windows 11 will increase drastically?
M.D.V.
If something has a solution... Why do we have to worry about?. If it has no solution... For what reason do we have to worry about?
Help me to understand what I'm saying, and I'll explain it better to you
Rating helpful answers is nice, but saying thanks can be even nicer.
|
|
|
|
|
sudo iexplore.exe sudo <do stuff>
|
|
|
|
|
The Caribbean island is reaping millions from .ai website registrations Time for the Bahamas to start cleaning up
OK, that "joke" requires you to do a bit of work and look up their domain name. Gotta keep the kid sisters safe!
|
|
|
|
|
It has been nearly 3 months since we launched Uno Platform 5.0. "One love, we get to share it"
|
|
|
|
|
A new report from cloud security company Sysdig reveals that many businesses are indulging in the dangerous practice of putting convenience before preventive security in pursuit of faster application development. Should I have warned you before posting this shocking news?
|
|
|
|
|
Must be a slow news day. This has long been known.
|
|
|
|
|
Both are true.
TTFN - Kent
|
|
|
|
|
A new Germany-based startup, Semron, is developing what it describes as “3D-scaled” chips to run AI models locally on smartphones, earbuds, VR headsets and other mobile devices. The article doesn't answer the important question: can they run DOOM?
|
|
|
|
|
Kent Sharkey wrote: ...can they run DOOM?
Not intelligently.
|
|
|
|
|
LLM on expandable RISC-like architecture sounds like a big win to me, but also very very hard to do.
edit: After rtfa, that's not exactly what they are shooting for, but close, and if someone does manage it, they will make bank. FPGAs could also be a similar route to the same result(s). The idea being hardware that can alter itself to model use-case to give performant operations against a specific model(s) (model-on-chips).
|
|
|
|
|
jochance wrote: hardware that can alter itself Oh, that's not scary at all, is it, Sarah Conner?
There are no solutions, only trade-offs. - Thomas Sowell
A day can really slip by when you're deliberately avoiding what you're supposed to do. - Calvin (Bill Watterson, Calvin & Hobbes)
|
|
|
|
|
FPGAs have been around a minute, they're pretty good for a certain class of computing problems.
Sure, pretty much everything about AI is scary. It's at least a good thing none of it is real. Most everything being called AI is actually ML.
Nothing resembling artificial general intelligence is probably even on the horizon. It won't stop people from thinking otherwise. It also won't stop "unintelligent" killer robots. But even ChatGPT and Bard are alive, depending who you ask. Not their fault or anything... Dr Sbaitso seemed pretty alive something like three decades ago too.
|
|
|
|
|
This post provides a behind-the-scenes look at how we test MSVC‘s backend, which is responsible for optimization and code generation. For those who doubted
|
|
|
|
|
Simulated 1-bit, 32x48 cellular grid runs at a blazing 0.00003 fps. Sure, teach the bacteria to run and gun. What could go wrong?
Although "run" seems way too generous for this bit of headlinery
|
|
|
|
|
Reading the article, the bacteria didn't run Doom. They display it, although very slowly and poorly.
|
|
|
|
|
Reading the article, I didn't write the title.
TTFN - Kent
|
|
|
|
|
Unprivileged attackers can get root access on multiple major Linux distributions in default configurations by exploiting a newly disclosed local privilege escalation (LPE) vulnerability in the GNU C Library (glibc). Good thing no one uses that library!
Even safer, as it's the error logging that leads to the hack. Don't log errors, and you're safe!
|
|
|
|
|
Kent Sharkey wrote: Even safer, as it's the error logging that leads to the hack. Don't log errors, and you're safe! Log? what's that?
M.D.V.
If something has a solution... Why do we have to worry about?. If it has no solution... For what reason do we have to worry about?
Help me to understand what I'm saying, and I'll explain it better to you
Rating helpful answers is nice, but saying thanks can be even nicer.
|
|
|
|
|
Microsoft's adoption of Rust continues apace if a posting on the company's careers website is anything to go by. If only they had R#
|
|
|
|
|
Like the recent article about compiling Rust to .NET IL?
It generally won't be "pretty" code, but going from IL to C# isn't terribly hard.
|
|
|
|
|
I bet this causes a panic!
All it means is that they will be rewriting some cloud services in Rust for performance, reliability and memory, similar to AWS. Clients will still be C# and polyglot.
Their quantum computing toolkit, Q#, has also just been rewritten in Rust from .NET. Though they haven't changed the name, so the unsuspecting might still think it's .NET based.
Kevin
|
|
|
|
|
Microsoft and OpenAI told Engadget the technique in question didn’t bypass their safety filters. Ixnay onyay ethay aiyay apocalypseyay
|
|
|
|
|
I've only really played with the stable diffusion stuff and running it on a desktop.
Depending on exactly where you grab what, it either has no real guardrails or has very very easily removable ones (change a line of code sort of deal).
I think this cat's out of the bag, effectively. I'm not sure it matters if malicious image generation comes from DALL-E or comes from StableDiffusion, or Midjourney, wherever.
The core issue is people disseminating them. Even if they were just really great handmade traditional digital art, this sort of thing shouldn't be considered "acceptable use".
Acceptable use - of the systems where such content is disseminated.
|
|
|
|