Examples
The examples below are real customer cases where ByteRouter was used to establish cost-efficient data analysis infrastructure, realising significant cost savings.
Molecular Docking of Antibodies
Client: Raven Biosciences
Large-scale molecular docking using HADDOCK software for signal protein analysis. Containerised applications handled data preparation, docking, and post-processing stages.
- Input: small files describing protein structures
- Output: 50–200 GB of intermediate data per run; minimal data required downstream
- Storage: S3-compatible object storage via DigitalOcean (US East)
- Infrastructure: Contabo bare-metal servers + existing on-premise server
- Processing volume: 5,000+ protein pairs analysed over 18 months
- Node utilisation: >80%
ByteRouter Cost
USD 8,855
AWS Equivalent
USD 50,301
Total Savings
USD 41,446
GPU-Accelerated Molecular Dynamics
Client: Raven Biosciences
Protein-protein interaction simulation using GROMACS software for binding affinity estimation of antibodies with protein targets. Performance measured in nanoseconds per day (ns/day).
- Simulation duration: several days per run
- Output: 1–20 GB per simulation
- Infrastructure: two geographically distributed data centres (Netherlands + US East Coast) plus an on-premise GPU server
- Total nodes: 18 GPU nodes
- Operational period: 8 months
- Average utilisation: >80%
ByteRouter Cost
USD 21,539
AWS Equivalent
USD 95,162
Total Savings
USD 73,623
Could ByteRouter work for your workload?
Both case studies involved small teams with no dedicated HPC expertise. If you have a compute-intensive pipeline, we can help.