We detect optimization potentials in GPU focused applications, for real-time and energy efficient data flows.
Measuring the energy efficiency of your software at code level, you have full control over your energy costs. identify the causes of high energy expenditure and take targeted countermeasures.
Don't let the code be the bottleneck in time critical simulations and get ahead of the flood of data.
Identify compute intensive calculations by measuring hot spots in your code. Knowing the bottlenecks, usage of hardware resources can be improved.
Customers who have already benefited from our performance analyses
Measure & visualize
Performance and energy measurements are done using open source tools. In addition we provide plugins for IDEs, HPC clusters scheduling tools. Our interactive dashboards will give you easy access to optimization potentials and estimation of cost reductions.
What is FLØW?
Fløw are a set of tools and methods to assess performance and energy efficiency of your codes. The main focus relies on compute intensive GPU applications ranging from conventional computer simulations to AI driven applications.
Why would I need Fløw?
Inefficient code means higher computing costs in terms of hardware usage, energy consumption and time
What is different to tool X, Y, Z?
We developed the open source tool EMA (Energy Measurements of Applications). In contrast to other solutions EMA not just measures the total energy costs of your application. It can be used to identify energy inefficient regions in your source code.
Can I use the tools in the HPC context?
We develop plugins for popular ressource schedulars like SLURM for the usage in High-Performance Computing.