A Teaspoon of HPC
Bringing computing and HPC to the masses is a quest that drives me. As England reaches a roasting 40 degrees and I struggle to type this, I could not be more aware of the climate crisis that affects my children and future generations around the world. We need to change almost every aspect of human systems, and that seismic change will not happen without an explosion of innovation. It is now well understood that diversity makes the best teams and that humans form new ideas building upon past experiences, so how do we bring computing and HPC to a wider range of people with a wider range of life experiences?
I recently attended the launch of the Raspberry Pi Computing Education Research Centre at Cambridge University, where I learnt about TSP (pronounced ‘teaspoon’) languages. TSP stands for task-specific programming. This concept comes from Professor Mark Guzdial of the University of Michigan, based not far from Altair headquarters. His quest is to bring a teaspoon of computing to every class through the use of TSP languages. In this way, students are exposed to computing concepts throughout their education and subject areas such as history, politics, and geography benefit from computer science. TSP languages do not have to be Turing-complete, but they do need to be usable and useful. They build upon ideas such as the visual code blocks of Scratch or even upon older concepts such as spreadsheet macros.
Why do other subject areas need computer science? Data manipulation and data visualisation are not just reserved for data scientists. My daughter’s school published data on Covid infections by class earlier in the pandemic. Historians need to be able to understand and explain population trends. Data science is not the same as computer science and does not have the same skill set. It is vital that we can enable one without the other. At Altair we offer low-code data science tools that enable people from many walks of life to adopt data science and gain powerful insight into their data.
How can we make HPC accessible?
There are many engineering and science disciplines that would benefit from a teaspoon of HPC to accelerate time to results, or simply to manage shared compute resources more effectively. Users should be able to run their applications on a supercomputer without needing to know how that computer works. They should have the same experience as if they are running on their own desktop. Altair® Access™ provides a simple, powerful, and consistent interface for submitting and monitoring jobs on compute clusters and cloud. By using application templates, a desktop user interface, and a smart data management layer, engineers and physicists can submit workloads without having to know about HPC. It allows engineers and researchers to focus on core activities and spend less time learning how to run applications and moving data around.
Adding a teaspoon of cloud
Supercomputers are breaking out of the datacentre and into the public cloud. With many public cloud vendors supporting powerful, memory-heavy machines with high-speed networking, it is now practical to leverage cloud for short-lived projects or bursty workloads. But HPC system administrators are busy people who don’t have the time to develop the means to extend the data into the cloud. That is where Altair cloud busting tools such as Altair® Control™ come in. They let you automatically spin up resources in the cloud when there are jobs waiting. They extend the HPC environment into the cloud and then take down the cloud machines when the workloads have finished. These powerful tools are used to orchestrate thousands of customer projects, scaling up and down dozens of clusters at a time, but they don’t have to work on that scale. The first step is always to take a teaspoon of cloud: one workload running on one or two machines. By setting limits and budgets in Altair cloud solutions, administrators can relax knowing that costs are bounded. Once comfortable with cloud, administrators can experiment with more sophisticated policies and automations, but that comes later once the first teaspoon has gone down well.
Looking to the future
We need to learn how to support humanity in a more sustainable way if we are going to save any aspect of the planet Earth we know today. Access to HPC and computing resources will play a central role in that though climate modelling, product design, and data science. Access to computing is also important for the individuals involved. Computing education is a powerful tool for social mobility, and it will to create the generation of people we need to solve the problems in the world today. So we need to support organisations such as the Raspberry Pi Foundation and centres for computing education so that students get access to the problem-solving and digital skills needed to be great engineers. At Altair we continue to focus on the convergence of simulation, HPC, and data science and how we can bring access to more people to change the world for the better.
About the author
Dr Rosemary Francis founded Ellexus, the I/O profiling company, in 2010, and Ellexus was acquired by Altair in 2020. Rosemary obtained her PhD in computer architecture from the University of Cambridge and worked in the semiconductor industry before founding Ellexus. She is now chief scientist for HPC at Altair, responsible for the future roadmap of workload managers Altair® PBS Professional® and Altair® Grid Engine®. She also continues to manage I/O profiling tools and is shaping analytics and reporting solutions across Altair’s HPC portfolio. Rosemary is a member of the Raspberry Pi Foundation, an educational charity that promotes access to technology education and digital making. She has two small children and is a keen gardener and windsurfer.