Our VP of Engineering, Pat McGarry, recently discussed how organizations should focus on putting their business needs and questions first, rather than compromising based on the limitations of their current technology. In the article, Pat breaks down how to gain business-critical insights and analysis from the crushing volume of big data into two components.

High Performance Computing Architectures
By using a heterogeneous computing model, organizations are provided with the right tool, at the right time, for the right job. Some processors and architectures are better at some things than others, similar to tools in your toolbox. Relying on a single processor type does not provide the most efficient tool or the performance and scalability needed to meet the demands of increasing high-velocity data.

FPGAs, CPUs, GPUS
Business Needs and Ease of Use
One of the complexities associated with heterogeneous computing is how to manage disparate computing resources. This can be achieved by providing open, compute-agnostic APIs that focus first on the business problem rather than implementing, designing and managing complex hybrid architectures. Allowing your organization to focus on the business problem and letting your data analytics technology work for you and for that problem is the ideal solution, with no requirements to understand the underlying technology to achieve needed business results.

Read more about how Ryft was able to achieve this and how data scientists can get back to focusing on the business problem when architectures are so complex in the full article in Scientific Computing Start with Business Needs, then Drive High Performance Computing Architectures. Below we have also included a video demonstration of the Ryft ONE performing the “Babe Ruth” fuzzy search example used in the article.

Hybrid (or Heterogeneous) Computing from Ryft on Vimeo.

Leave a Reply

Your email address will not be published. Required fields are marked *