kernel_tuner VS RSS-Link-Database

I've created Kernel Tuner (https://github.com/KernelTuner/kernel_tuner) as a small software development tool, because I was writing a lot of CUDA and OpenCL kernels at the time. I didn't want to manually figure out what best thread block dimensions and work division among threads were on every GPU over and over again.

The tool evolved quite a bit since the first versions. I'm also using it for testing GPU code, teaching, and it has become one of the main drivers behind a lot of the research that I do.

We have an open science policy, so anyone can use our framework yourself to optimize stuff, if you want! The original paper is linked at the bottom of the GitHub page.

This is a great post for people who are new to optimizing GPU code.

It is interesting to see that the author got this far without interchanging the innermost loop over k to the outermost loop, as is done in CUTLASS (https://github.com/NVIDIA/cutlass).

As you can see in this blog post the code ends up with a lot of compile-time constants (e.g. BLOCKSIZE, BM, BN, BK, TM, TN) one way to optimize this code further is to use an auto-tuner to find the optimal value for all of these parameters for your GPU and problem size, for example Kernel Tuner (https://github.com/KernelTuner/kernel_tuner)

On the other hand it is not 1995. Time has moved on. I wrote a Simple RSS feed, that also serves as search engine for bookmarks.

I am able to run it in attick on raspberry pi. We do not have to rely so heavily on google.

https://github.com/rumca-js/Django-link-archive

It is true that it does not serve me as google, or kagi replacement. It is a very nice addition though.

With a little bit off determination I do not have to be so dependent on google.

Here is also a dump of known domains. Some are personal.

https://github.com/rumca-js/Internet-Places-Database

...and my bookmarks

https://github.com/rumca-js/RSS-Link-Database

Some more years, and google can go to hell.

[4] https://github.com/rumca-js/Django-link-archive

These are exported then to github repositories:

[5] https://github.com/rumca-js/RSS-Link-Database - bookmarks

[6] https://github.com/rumca-js/RSS-Link-Database-2023 - 2023 year news headlines

[7] https://github.com/rumca-js/Internet-Places-Database - all known to me domains, and RSS feeds

My own repositories:

- bookmarked entries https://github.com/rumca-js/RSS-Link-Database

- mostly domains https://github.com/rumca-js/Internet-Places-Database

- all 'news' from 2023 https://github.com/rumca-js/RSS-Link-Database-2023

I am using my own Django program to capture and manage links https://github.com/rumca-js/Django-link-archive.

That is why I wrote [1] for myself. It stores links in database, which I can query. Everything is later on exported, like in [2] and [3]. I can browse history, I can find useful data. I do not say it has replaced google for me. It is a nice addition that helped me gather data I encounter on the Internet.

It is a link database, at first glance resembles Reddit clone, but my focus is on creating link database, not on providing social media experience cancer.

Links:

[1] https://github.com/rumca-js/Django-link-archive

[2] https://github.com/rumca-js/RSS-Link-Database

[3] https://github.com/rumca-js/RSS-Link-Database-2023

- https://github.com/rumca-js/RSS-Link-Database-2022 - all captured links in 2022

- I think newsboat reader does not provide advanced search mechanism

There is also a repo in which I store entries which I find interesting, useful: https://github.com/rumca-js/RSS-Link-Database