Sign up for our newsletter and get the latest HPC news and analysis.
Send me information from insideHPC:


Parallel Computing in Python: Current State and Recent Advances

Pierre Glaser from INRIA gave this talk at EuroPython 2019. “Modern hardware is multi-core. It is crucial for Python to provide high-performance parallelism. This talk will expose to both data-scientists and library developers the current state of affairs and the recent advances for parallel computing with Python. The goal is to help practitioners and developers to make better decisions on this matter.”

Video: High-Performance Computing with Python – Reducing Bottlenecks

This course addresses scientists with a working knowledge of NumPy who wish to explore the productivity gains made possible by Python for HPC. “We will show how Python can be used on parallel architectures and how to optimize critical parts of the kernel using various tools. The following topics will be covered: – Interactive parallel programming with IPython – Profiling and optimization – High-performance NumPy – Just-in-time compilation with Numba – Distributed-memory parallel programming with Python and MPI – Bindings to other programming languages and HPC libraries – Interfaces to GPUs.”

NVIDIA DGX-2 Delivers Record Performance on STAC-A3 Benchmark

Today NVIDIA announced record performance on STAC-A3, the financial services industry benchmark suite for backtesting trading algorithms to determine how strategies would have performed on historical data. “Using an NVIDIA DGX-2 system running accelerated Python libraries, NVIDIA shattered several previous STAC-A3 benchmark results, in one case running 20 million simulations on a basket of 50 instruments in the prescribed 60-minute test period versus the previous record of 3,200 simulations.”

Making Python Fly: Accelerate Performance Without Recoding

Developers are increasingly besieged by the big data deluge. Intel Distribution for Python uses tried-and-true libraries like the Intel Math Kernel Library (Intel MKL)and the Intel Data Analytics Acceleration Library to make Python code scream right out of the box – no recoding required. Intel highlights some of the benefits dev teams can expect in this sponsored post.

Intel High-Performance Python Extends to Machine Learning and Data Analytics

One of the big surprises of the past few years has been the spectacular rise in the use of Python* in high-performance computing applications. With the latest releases of Intel® Distribution for Python, included in Intel® Parallel Studio XE 2019, the numerical and scientific computing capabilities of high-performance Python now extends to machine learning and data analytics.

Python Power: Intel SDK Accelerates Python Development and Execution

It was with one goal – accelerating Python execution performance – that lead to the creation of Intel Distribution for Python, a set of tools designed to provide Python application performance right out of the box, usually with no code changes required. This sponsored post from Intel highlights how Intel SDK can enhance Python development and execution, as Python continues to grow in popularity.

HiPEAC Vision 2019 Looks to the Future of Computing

“Today, the possibilities of an interconnected, heterogeneous and intelligent world are only just beginning to make themselves known. This stunning advancement in digital technology was made possible by ever-increasing performance at ever lower costs. However, physical limits mean we won’t be able to keep shrinking computing components while increasing performance for much longer. So where do we go from here? What are the main challenges and conditions for future developments, and where? The HiPEAC Vision 2019 explores all these questions, and more.”

Intel Performance Libraries Accelerate Python Performance for HPC and Data Science

Python is now the most popular programming language, according to IEEE Spectrum’s fifth annual interactive ranking of programming languages, ahead of C++ and C. Recent Intel Distributions for Python show that real HPC performance can be achieved with compilers and library packages optimized for the latest Intel architectures. Moreover, the library packages targeted for big data analysis and numerical computation included in this distribution now support scaling for multi-core and many-core processors as well as distributed cluster and cloud infrastructures.

Machine Learning with Python: Distributed Training and Data Resources on Blue Waters

Aaron Saxton from NCSA gave this talk at the Blue Waters Symposium. “Blue Waters currently supports TensorFlow 1.3, PyTorch 0.3.0 and we hope to support CNTK and Horovod in the near future. This tutorial will go over the minimum ingredients needed to do distributed training on Blue Waters with these packages. What’s more, we also maintain an ImageNet data set to help researchers get started training CNN models. I will review the process by which a user can get access to this data set.”

Python Can Do It

“Python remains a single threaded environment with the global interpreter lock as the main bottleneck. Threads must wait for other threads to complete before starting to do their assigned work. The result of this model is that production code is produced that is too slow to be useful for large simulations.”