Monthly Archives: August 2024

Eclipse – too small icons on high dpi screens – workaround for KDE Plasma by scaling on X11 and Wayland

Posted on by

I have to work with Eclipse (version 4.32) again for a while. After two days of relatively intensive work, I am a bit frustrated. One reason is the small size of the icons – 16×16 px. Not only in the main icon taskbar. There are minuscule symbols elsewhere – e.g. in the bar on the left side of the editor section. Regarding e.g. the tiny symbol for code folding: You can almost not see the “+” or “-” symbols on a high resolution screen …

One can scale almost all font-sizes within Eclipse – but not such a basic thing as the icon/symbol size or the icon bar height.

I have three QHD screens (2560×1440) at my present workplace. At another site there are 2 4K screens. The following image can give you an impression for the QHD situation – if you have a QHD, UHD or 4K screen yourself and not zoomed the browser contents already. For screens with lower resolution see below.

Before you say this looks quite OK, check the following:

  • Open this post on a desktop screen in a window wider than 800 px.
  • Enforce normal content size size in your browser (no zoom! Ctrl-0).
  • Get a natural distance to your working QHD, UHD or 4K monitor (like 80+ cm). In case you have a QHD screen you get already the right impression.
  • Users who just have a monitor with a 1920×1200 resolution must in addition downscale their browser contents to around 80% (QHD) or 55% (UHD, 4K) to get the right impression of the problem. (After Ctrl-0)
  • Folks with QHD get the right 4K impression by scaling down to 66%. (After Ctrl-0)

While for QHD the working conditions may still appear to be OK, you get a real problem for UHD and 4K resolutions. The font-size may still be OK (after having adapted it). The unadaptable symbol sizes, however, are far too small. At least in my opinion. But even on my standard QHD screens I personally do not feel comfortable after a while.

In this post I want to share some workarounds. The methods described may be interesting for other GTK applications, too. Keep in mind that the images in this post may give you a wrong impression if you zoomed the browser contents. But later on, when I demonstrate the effect of some measures, only relative sizes of fonts and icons are of interest.

Continue reading →

Eclipse IDE – how to limit the number of of CPU cores used on a Linux PC

Posted on by

In the very hot days which we experienced in Germany the last week my old ventilation cooled PC suffers a bit when all of my CPU cores get under full load. Temperatures up to 70&grad; Celsius despite an Alpenföhn ventilator are beyond my normal experience and tolerance. The culprit was Eclipse with PyDev.

I had installed the latest Eclipse version (4.32.0) some days before. And opened a new Python project based on a directory with tons of code for Machine Learning classes. As a default the PyDev validator started to check the code for problems. It does so by using all available CPU cores and threads in parallel. Yeah, Eclipse works highly parallelized with its underlying Java VM. As I had a lot of complex code the code analysis lasted for some minutes – putting an extreme load on my PC.

So, my question was: How can one limit the number of CPU cores/threads used by Eclipse?

For Linux the answer is surprisingly simple:

We can use the “taskset” command. E.g., if I wanted to use the first 4 CPU cores (or recognized threads) we would write on the command line: 

taskset -c 0-3 ./eclipse &

This works perfectly! The initial code analysis takes a bit longer – but the load on the PC got down. Happy coding!

 

Machine Learning on PCs – Use mixed precision and look out for super-convergence to save energy

Posted on by

People doing Machine Learning [ML] experiments on their own Linux PCs or laptops know that the numerical training runs put a heavy load on the graphics cards and consume a lot of energy as a direct consequence. Especially in a hot summer like we have it in Germany right now, cooling of your systems may become a problem. And as energy has a high price tag here, any method to reduce the load and/or power consumption is welcome.

But I think that caring about energy consumption is a topic which we as a Linux and ML enthusiasts should keep in mind in general. Some big tech companies will probably not do it – as long as their money machinery works and as some heads follow fantasies about building small nuclear power plants for their big AI data centers. But we Opensource people would like to see more AI- and ML-services independent of the monopolists and their infrastructure, anyway. Not only for reasons of data and privacy protection.

As soon as we, however, proclaim and work for a development that favors local and resource optimized installations of AI and ML tools both for private people and companies, we have to care about side effects: We have to bring the energy consumption down for these many local installations substantially in parallel. Otherwise, centralized solutions may have a better energy efficiency than decentralized solutions.

For me as a retired person in Germany the general financial pressure is high enough to enforce a careful use of my private resources. With this post I want to draw your attention to two points which may help you, too, to save energy during your ML-experiments. (In addition to or aside of standard measures like saving certain model states during training runs to get better starting points for new runs.)

Continue reading →