Ellexus: What was great in 2019

It’s been an amazing year for Ellexus. The theme for 2019 has been preparation and discovery: preparing workloads for hybrid cloud, preparing on-prem systems for exascale, preparing for new storage technologies, discovering legacy applications and discovering areas for optimisation and opportunities for change. In light of this I expect to see the snowball continue as the hybrid cloud offerings for HPC become more sophisticated, we move closer to exascale and the industry comes out with ever more complex storage solutions.

Thank you to all our customers, partners and friends who’ve worked with us to make it a really great year. Here are just a few highlights in case you missed them. Happy Holidays all and I look forward to 2020!

The year that was 2019

At the start of 2019 Rosemary travelled to Texas to present at the Rice Oil and Gas workshop on HPC. Rosemary was invited to speak about how to benchmark workloads and profile I/O when preparing for Cloud and Exascale.

Still in the US, Ellexus also worked with HPC pros, Sylabs Singularity and Rescale to run a DAC workshop at the DAC conference for the semiconductor industry. We were showcasing how Qualcomm runs Breeze as a Service to deliver automatic lists of dependencies for workloads, a critical part of migrating and containerising workloads. The focus of the workshop was how to make hybrid cloud work for EDA workloads.

In Europe there were opportunities to show off the UK’s HPC excellence. Rosemary presented at the Spectrum Scale User Group at ISC on the benefits on running our I/O monitoring technology at the UK’s Diamond Light Source. Also in the UK, we worked with the STFC Hartree Centre to further the UK’s position of excellence in HPC, improving performance by more than 8%.

We continue to work with the Sanger Institute to optimise genome pipelines that will change the lives of so many of us. The work they have been doing to make their pipelines faster and more portable opens up the market to both more efficient on-prem usage as well as hybrid cloud models.

Back on the other side of the pond we collaborated with Sandia National Lab on preparing the IO500 benchmark for their Aarch64 Arm cluster. Supercomputing had barely finished before we headed to Vegas for AWS Re:Invent to see Arm present on their journey towards using data science to understand workloads and plan for the future: truly the technology of tomorrow.