Share this:


I have mentioned before that one of my favorite parts of my position here at Nimbix is that I get to interface with all of our customers and prospective customers. This means I get to see all of the amazing ways they are using technology to solve problems, implementing creative new business ideas and models.

One such model, I have seen more and more of late is where a user leverages our “On-Demand” heterogeneous HPC cluster as the processing engine for their solution. This allows them to have very little to no standing infrastructure cost. Often they may only have a web and/or database server that handles their interactions and transactions with their clients, but all of the expensive High Performance Processing is being handled by Nimbix’s NACC (Nimbix Accelerated Compute Cloud) in an on-demand basis.

This is possible because of a few reasons:

  • Nimbix offers the ability to run “Privatized” applications in our NACC and only allows specific authorized users to run them.
  • NACC is batch workload/job based.
  • Workloads/Jobs can be submitted to NACC via a simple API call.
  • NACC is always listening for requests and monitoring available resources to fill them.
  • NACC’s ability to scale. Workloads can be launched concurrently on separate nodes reducing wait times.
  • Nimbix only charges for the actual processing time used.
  • Processing times are billed to the minute.

What makes this business model so interesting is how it creates a very efficient cost structure. For every workload the customer sends to Nimbix for processing they know they have a paying customer on their end to support it. This means there is no cost associated with having expensive specialized hardware waiting around for work. This is an ideal cost model and environment for startups or businesses with unpredictable levels of demand that require High Performance Computing resources. It turns the expensive fixed cost normally associated with High Performance Computing into a variable cost that ebbs and flows with their business.

Of course, the other obvious benefit this offers is scalability. For example if the customer gets 5 jobs at the same time they can run them concurrently on 5 different servers in NACC and get the results back to their customers faster than having to queue them up to run through a single dedicated server. Or if they have one large job they can break it up into a few smaller ones and run those across several nodes as well. Basically, it delivers on the promise of true elasticity we have been hearing about for so long.

Of course there is a tipping point when there is enough steady business that it cost justifies setting up a dedicated server or cluster. The beauty of NACC is it can be used in conjunction with the dedicated environment to provide that same elasticity for periods of peak demand.

For example lets say the dedicated servers are busy with a large job(s) that will monopolize them for several hours, but in the meantime the customer has received several small jobs from some of their other customers. They don’t want to have their customers wait several hours to get their results back on these small jobs or they run the risk of them having a bad experience and losing any future business from those customers.

By leveraging NACC those smaller jobs can be submitted to Nimbix to be processed quickly, while their dedicated servers continue to work on the larger job(s).  This way all of the customers are happy and keep coming back.

One other important point to note is that obviously, the dedicated and front end servers can be hosted anywhere, but for the best possible performance Nimbix can host dedicated environments in the same datacenter as our NACC making data transfer that much faster and easier.

The flexibility that Nimbix provides allows this highly efficient model to grow with our customers business needs there by providing  an optimal cost model and an easy path to profitability.  Please, contact us if you are curious about how we can help meet your specific needs.  Contact Us