Embedded Files
Cost Effective High Performance Computing
Cost Effective High Performance Computing
ByteRouter is a product for establishing cost efficient infrastructure and using that infrastructure for high performance computing workloads such as data analysis and simulations through a user-friendly interface.
10x cost reductions
10x cost reductions
An example cost per CPU core for different hosting options with comparable performance. Although this comparison is simplified, it shows that significant savings can be realized by choosing the right infrastructure.
*On-premise: Hardware you own and manage
**Datacenter: Rented hardware in a professional datacenter
The Challenge
The cloud makes it easy to deploy, scale and maintain complex infrastructure without the need for capital expenditure. However, this convenience comes at a cost and for compute intensive workloads, significant savings can be realized by using more cost-efficient solutions.
Today, professional data centers offer enterprise grade hardware for a fraction of the cost of comparable cloud services. Customers can rent or lease infrastructure managed by the vendor and avoid both capital expenditure and operational costs. However, scaling such infrastructure for complex data analysis pipelines is often a complex task that can erode any cost savings.
A Solution
ByteRouter is a software product designed for running complex data analysis pipelines on any hardware available, on-premise, hosted or cloud. The necessary infrastructure is defined through a user-friendly interface and exposed as a virtual cluster to the users.
The integrated scheduler and data management layer, allow complex data analysis pipelines to be executed across multiple datacenters to leverage the most cost-efficient hardware. ByteRouter is also focused on simplifying maintenance of complex clusters to keep operational costs low and achieve maximum cost savings.
Use Cases
Use Cases
Teams working with computationally demanding data analysis such as data science, AI or bioinformatics can use ByteRouter to establish a cost-effective analysis infrastructure by renting hardware at vendors specializing in bare metal services (e.g. Contabo, HOSTKEY or Hetzner) and integrate with cloud services or on-premise hardware as needed.
Engineering simulation that requires High Performance Computing (e.g. ANSYS or OpenFOAM) can use ByteRouter as a light-weight alternative to Slurm to quickly establish and right-size a cluster for heavy simulations on cost-effective hardware.
Molecular simulation applications such as Molecular Docking (e.g. Glide, HADDOCK) or Molecular Dynamics (e.g. GROMACS, LAMMPS) are also applications that require expensive infrastructure to scale well. ByteRouter can be used to quickly give access to both CPU and GPU nodes.
See the examples section for detailed examples of how ByteRouter has been deployed for customers.
Fill out the contact form if you want to get a demo of ByteRouter or discuss how we can help you establish cost-efficient HPC infrastructure
Page updated
Google Sites
Report abuse