Based on previous experience, this winter, we’ll be hearing about storms heading our way from the Atlantic. Maybe you are thinking of buying a new car this year and want to check out its safety record? Alternatively, you might be interested in advances in the diagnosis of genetic diseases?
Compute-intensive tasks
All these seemingly different scenarios have something in common! Yes – weather modelling, car crash simulation testing and genome-sequencing are all compute-intensive tasks that need huge amounts of processing power. Other vertical industries include defence and security, geographic information engineering, financial high frequency trading plus oil and gas exploration. These are all areas where High Performance Computer Clusters (HPCs) come into their own.
High-cost and locked-in
Traditionally, HPCs were built using high-cost, proprietary hardware platforms, designed specifically for a single use case. As a result, they were the preserve of very large organisations that could afford to invest millions of dollars. They were costly to design, maintain and upgrade, and of course, different HPCs – using different protocols – typically didn’t speak to one another. What’s more, the traditional HPC solution was delivered in component parts rather than an integrated, fully tested appliance. This added cost, time and complexity.
Open standards
Fast forward to today. The industry is transforming, thanks to open standards and non-proprietary hardware platforms, based on standard IT technology, pioneered by companies like Dell EMC. This has not only made HPCs affordable, it is also enabling increased use of ground-breaking applications like Artificial Intelligence, Machine Learning, genome-sequencing and Finance and KAFKA-ingest streaming. Increasingly, we are seeing Field Programmable Gate Arrays (FPGAs) being deployed across a range of on-premise platforms and cloud infrastructure to deliver a big step-change in application performance and energy-efficiency.
Collaboration
Exactly why I’m so excited that the division I head up, Dell EMC OEM, is now collaborating with BittWare, a leading supplier of FPGA-based accelerated computing solutions. Together, we are bringing Enterprise-class OEM solutions to the market, allowing customers to drive the most demanding data visualisation and rendering workloads.
Choice and flexibility
For me, it’s all about customer choice, cost and flexibility. You get a 1U high-density, rack-mounted PowerEdge server appliance solution that has been optimised for accelerators and includes the BittWare FPGA Compute Node. Forget about getting a bag of component parts! This is an integrated, customised, production-ready solution that has been tested and validated in our labs by our HPC team and backed up by award-winning support and a fast go-to-market strategy. Think affordable, robust, scalable technology from two top vendors. Information available here.
My counterpart, Craig Petrie, VP Marketing of FPGA Solutions at BittWare, a Molex company, hit the nail on the head, when he said, “Packaging disruptive technology in this way massively de-risks customer investments, allowing them to unlock the dramatic benefits of FPGAs in a cost-effective manner.” I couldn’t agree more!
I believe that Dell EMC OEM and BittWare are in a great position to lead the way in FPGA technology and disrupt the market for the benefit of customers. Are you using HPC technology? What are your thoughts? I’d love to hear your comments and questions.
Learn more about Dell EMC OEM at dellemc.com/oem
Learn more about our joint solution here: https://www.bittware.com/facn
Keep in touch. Follow us on Twitter @dellemcoem and @dermotatdell
Join our LinkedIn OEM & IoT Solutions Showcase page here