HPCs build super-computers, not networks, right? Well you’d think so, but that could be about to change as data generated by European research institutes and experiments continues to explode. With such rapid data growth, many HPCs are finding that traditional managed network services are beginning to restrict their ability to share data with research partners and clients in a timely way. As the sheer volumes of data continue to test networks, HPCs’ ability to monetize major investments in computing and storage infrastructure may also be impacted.

As the issue of growing data volumes becomes more acute, HPCs are looking for new ways to speed up the process of sharing data with partners and clients – which is where dark fibre comes in.

HPCs are looking for new ways to speed up the process of sharing data with partners and clients – which is where dark fibre comes in.

Change is always challenging, though, and the move from managed services to dark fibre is no exception. HPCs need concrete reasons to consider building their own networks – and here are our top 5:

1. More flexibility based on fast, agile connections - With dark fibre networks, you no longer wait for managed services partners to provision or scale connections for you. Instead, you control the equipment that performs these functions, so you can roll out new connections and turn-up new services in hours, not weeks, and scale services in real time to handle traffic peaks.
   
   
2. Future-proof networking as data grows - Optical and packet technologies exist that offer virtually unlimited bandwidth over dark fibre connections, future-proofing networks in the face of exploding data volumes. One example is Ciena’s Waveserver Ai, which was recently used by CESNET and GÉANT to deploy a 300G network for the research and education community in the Czech Republic. The project will enable CESNET to meet the rapidly growing bandwidth demands of Europe’s academic and scientific organizations for years to come.
   
   
3. Simple deployment and management - Mature optical and packet technologies make it fast and simple to light dark fibre on your NREN or other provider’s network, making this approach to network building a viable option for HPCs for the first time. Additionally, the leading network equipment providers offer robust, open, comprehensive management systems that centralize and simplify your network operations.
   
   
4. Efficient networking with reduced data-transport costs - The leading optical and packet networking technologies converge traffic from layers 0 to 2 in a single environment, helping to minimize your network footprint and power and cooling costs. Additionally, in-house deployment and management of network equipment offers significant cost savings compared to managed services, with total cost predictability.
   
   
5. The first step to an adaptive network - Innovative network providers offer a clear roadmap to help HPCs migrate to the adaptive networks of the future. With software that supports network functions of virtualization, orchestration and automation, HPCs will be able to minimize manual management interventions and optimize network utilization, reducing network costs and specialist skills requirements.

These five reasons show that building dark fibre networks is a sensible decision, not only from a cost-efficiency standpoint, but also in terms of future-proofing your operations and keeping pace with surging bandwidth demands. The ability to move high volumes of data from and to the HPC infrastructure is vital to a successful business operation, and that is exactly what a dark-fibre network enables. 

Added to that, the maturity of the industry’s leading optical and packet networking technologies means that lighting dark fibre is no longer a leap of faith– it’s a straightforward exercise that HPCs can take in their stride.

Get in touch for more information on how Ciena can help you make the shift from managed services to fast, agile, efficient dark fibre networks.