ericjmorey

Jul 1, 2024

Aman Salykov writes:

This blog post is the result of my attempt to implement high-performance matrix multiplication on CPU while keeping the code simple, portable and scalable. The implementation follows the BLIS design, works for arbitrary matrix sizes, and, when fine-tuned for an AMD Ryzen 7700 (8 cores), outperforms NumPy (=OpenBLAS), achieving over 1 TFLOPS of peak performance across a wide range of matrix sizes.

By efficiently parallelizing the code with just 3 lines of OpenMP directives, it’s both scalable and easy to understand. The implementation hasn’t been tested on other CPUs, so I would appreciate feedback on its performance on your hardware. Although the code is portable and targets Intel Core and AMD Zen CPUs with FMA3 and AVX instructions (i.e., all modern Intel Core and AMD Zen CPUs), please don’t expect peak performance without fine-tuning the hyperparameters, such as the number of threads, kernel, and block sizes, unless you are running it on a Ryzen 7700(X). Additionally, on some Intel CPUs, the OpenBLAS implementation might be notably faster due to AVX-512 instructions, which were intentionally omitted here to support a broader range of processors. Throughout this tutorial, we’ll implement matrix multiplication from scratch, learning how to optimize and parallelize C code using matrix multiplication as an example.

Read Beating NumPy's matrix multiplication in 150 lines of C code

Niklas Gray writes:

I often find that when I work on the low level implementation of something I discover ideas that I can bring back and use to inform the high level design — to make it easier to work with, more performant, more orthogonal, etc. Thus, the flow goes back and forth from high-level to low-level, instead of just in one direction. If I’m stuck in some part of the high-level design, starting to work on the implementation is often the best way to get unstuck.

One such idea, which is really simple, but tends to lead to better and simpler code, is the idea of defaulting to zero. I.e. to always use 0 as the default or nil value. For example...

Read Defaulting to Zero

cross-posted from: https://programming.dev/post/9979951

Monday, February 12, 2024
Ribbon and Ron Williams write:

As Redox functionality becomes more complete, we have been working hard to get a wide variety of software working.

This post will cover our porting strategy for Linux/BSD programs.

We have ported the following games and emulators:

2048
ClassiCube
DevilutionX
DOSBox
eduke32
FreeCiv
Gigalomania
Hematite
Mednafen
Neverball
OpenJK
OpenTTD
PrBoom (Doom engine)
ScummVM
Space Cadet Pinball
and others.

Porting is a major part of the Redox development effort. We are using porting as a way to prioritize and validate Redox functionality.

Currently dozens of programs and many more libraries work. Our initial focus has been on porting Rust programs, but we also recognize the importance of supporting programs written in other languages.

In last year Ribbon began the porting of more than 1000 programs and libraries to Redox! They are still work-in-progress and many require customized cross-compilation scripts or improved library support. You can see them here.

With our recent change to a Linux-compatible path format, we have removed a major hurdle to supporting Linux applications. In the future we plan to expand our POSIX support, port more Rust crates and continue to improve Relibc.

Some thought is being given to virtual machines and Wine as possible mechanisms for running proprietary binaries and possibly even proprietary drivers. However, there are no specific plans for that capability at this time.

Read Porting Strategy - Redox - Your Next(Gen) OS

If all goes well, we can release the alpha by the end of March

Now Available on NixOS!

COSMIC Terminal joins a long list of COSMIC DE projects that have been packaged for NixOS! While COSMIC DE will be the flagship desktop for Pop!_OS, it's designed and architected for distros to build their own custom, branded user experiences. Details for distro packaging will be released at the same time as the alpha.

cross-posted from: https://programming.dev/post/8149733

Andrew Cunningham (arstechnica.com) - Jan 4, 2024 8:01 am UTC Writes:

Microsoft pushed throughout 2023 to add generative AI capabilities to its software, even extending its new Copilot AI assistant to Windows 10 late last year. Now, those efforts to transform PCs at a software level is extending to the hardware: Microsoft is adding a dedicated Copilot key to PC keyboards, adjusting the standard Windows keyboard layout for the first time since the Windows key first appeared on its Natural Keyboard in 1994.

The Copilot key will, predictably, open up the Copilot generative AI assistant within Windows 10 and Windows 11. On an up-to-date Windows PC with Copilot enabled, you can currently do the same thing by pressing Windows + C. For PCs without Copilot enabled, including those that aren't signed into Microsoft accounts, the Copilot key will open Windows Search instead (though this is sort of redundant, since pressing the Windows key and then typing directly into the Start menu also activates the Search function).

A quick Microsoft demo video shows the Copilot key in between the cluster of arrow keys and the right Alt button, a place where many keyboards usually put a menu button, a right Ctrl key, another Windows key, or something similar. The exact positioning, and the key being replaced, may vary depending on the size and layout of the keyboard.

We asked Microsoft if a Copilot key would be required on OEM PCs going forward; the company told us that the key isn't mandatory now, but that it expects Copilot keys to be required on Windows 11 keyboards "over time." Microsoft often imposes some additional hardware requirements on major PC makers that sell Windows on their devices, beyond what is strictly necessary to run Windows itself.

Read Microsoft is adding a new key to PC keyboards for the first time since 1994

cross-posted from: https://programming.dev/post/8121843

~n (@nblr@chaos.social) writes:

This is fine...

"We observed that participants who had access to the AI assistant were more likely to introduce security vulnerabilities for the majority of programming tasks, yet were also more likely to rate their insecure answers as secure compared to those in our control group."

[Do Users Write More Insecure Code with AI Assistants?](https://arxiv.org/abs/2211.03622?

cross-posted from: https://programming.dev/post/8121669

Taggart (@mttaggart) writes:

Japan determines copyright doesn't apply to LLM/ML training data.

On a global scale, Japan’s move adds a twist to the regulation debate. Current discussions have focused on a “rogue nation” scenario where a less developed country might disregard a global framework to gain an advantage. But with Japan, we see a different dynamic. The world’s third-largest economy is saying it won’t hinder AI research and development. Plus, it’s prepared to leverage this new technology to compete directly with the West.

I am going to live in the sea.

www.biia.com/japan-goes-all-in-copyright-doesnt-apply-to-ai-training/

cross-posted from: https://programming.dev/post/8121669

Taggart (@mttaggart) writes:

Japan determines copyright doesn't apply to LLM/ML training data.

On a global scale, Japan’s move adds a twist to the regulation debate. Current discussions have focused on a “rogue nation” scenario where a less developed country might disregard a global framework to gain an advantage. But with Japan, we see a different dynamic. The world’s third-largest economy is saying it won’t hinder AI research and development. Plus, it’s prepared to leverage this new technology to compete directly with the West.

I am going to live in the sea.

www.biia.com/japan-goes-all-in-copyright-doesnt-apply-to-ai-training/

cross-posted from: https://programming.dev/post/8098358

KDE set up their own Lemmy instance and has an active KDE community [relative link]

cross-posted from: https://programming.dev/post/8098358

KDE set up their own Lemmy instance and has an active KDE community [relative link]

KDE set up their own Lemmy instance and has an active KDE community [relative link]