Python 3.13.0 built with LLVM 19 and full LTO

Python 3.13.0 built with LLVM 19 and full LTO

Finally! We managed to compile the code of Python 3.13.0 using LLVM 19, with full link-time optimization (LTO): The installation passes all 44 standard optimization tests: Using random seed: 4347316840:00:00 load avg: 1.72 Run 44 tests sequentially in a single process0:00:00 load avg: 1.72 [ 1/44] test_array0:00:00 load avg: 1.72 [ 2/44] test_base640:00:00 load avg: 1.72 [ 3/44] test_binascii0:00:00 load avg: 1.72 [ 4/44] test_binop0:00:00 load avg: 1.72 [ 5/44] test_bisect0:00:00 load avg: 1.72 [ 6/44] test_bytes0:00:03 load avg: 1.67…

Read More Read More

Towards full adoption of LLVM: building a productive SQLite3 library installation required by TensorFlow

Towards full adoption of LLVM: building a productive SQLite3 library installation required by TensorFlow

Nowadays, the management of TensorFlow models using the SQLite3 library is becoming more and more demanded. It has become a standard practice for applicants for HPC resources to insist on accessing an optimized installation of the SQLite3 library, and avoid the one provided by the Linux distribution or delivered by Conda. Until one becomes involved in dealing with the constant changing of large sets of models, that might seem like an exaggeration or overstatement. It should be noted that the…

Read More Read More

Discoverer is advancing its commitment towards the full adoption of the LLVM compiler infrastructure

Discoverer is advancing its commitment towards the full adoption of the LLVM compiler infrastructure

The new version of the LLVM Compiler Infrastructure, 19, is available in the software repository. That version includes numerous improvements and fixes. We uphold our commitment to maintain openness and transparency in our way of doing system and software engineering, which entails consistently publishing our build recipes. In the case of LLVM 19, those recipes are publicly available here: https://gitlab.discoverer.bg/vkolev/recipes/-/tree/main/llvm/19 Adopting the LLVM 19 puts a new milestone in our computation model. We can start confidently moving the entire executables…

Read More Read More

[Computed on Discoverer] Calculation and visualization of protein conformations generation in implicit solvent

[Computed on Discoverer] Calculation and visualization of protein conformations generation in implicit solvent

We have numerous results we would like to share with you. One instance is the computation that demonstrates the conformation changes that make the protein to fold and unfold. The proper analysis of those changes help to measure the free energy of different processes related to the protein chain transformations. Our protein model representation depicts the side chains of each residue by utilizing a coarse grain force field, as illustrated in the video slides. By doing this, we significantly reduce…

Read More Read More

[Computed on Discoverer] Computation and visualization of ensemble averages

[Computed on Discoverer] Computation and visualization of ensemble averages

The molecular dynamics simulations provide great details at the nanoscale level for relatively big systems (up to several million atoms). Very often the material environment of certain functional groups, selected atoms, clusters of atoms, and other low to high-level structural elements, is numerically evaluated employing radial distribution functions (RDF). That method is powerful, well studied, and largely implemented for analysis. But it cannot provide a visual spatial estimate for the investigated environment. If you have a molecule surrounded by ions,…

Read More Read More

As we bid farewell to 2023…

As we bid farewell to 2023…

Now, as the year 2023 draws to a close, it is time to reflect upon the closest past. During the year 2023, we at Discoverer faced many challenging technical issues and took immediate action when it was needed, sometimes sacrificing our personal and family spare time to do so. However, ultimately, we accomplished all of our tasks. The cluster performs satisfactorily, and the majority of customers are content with the efficiency of the computational resources. Santa Claus brought us many…

Read More Read More

Zlib 1.3 has landed on Discoverer

Zlib 1.3 has landed on Discoverer

To support the compression/decompression abilities embedded in various software products that run on Discoverer, our team compiled and installed an AMD EPIC-optimized version of the Zlib 1.3 library. It is available in several variants, depending on the compiler used to compile the code, to support the benchmarking and development projects. Intel oneAPI and NVIDIA HPC SDK specific libraries are statically compiled into libz.so. Both AMD AOCC and LLVM build infrastructure LLVM builds are available. GCC build is also installed and…

Read More Read More

WordPress Appliance - Powered by TurnKey Linux