Non Stop Trends & Wins

NonStop – HPC and Exascale

High-Performance Computing (HPC) also known as the technology required for Artificial Intelligence (AI), Machine Learning (ML) and Deep Learning (DL) is getting even better.  I have a great interest in this and if you’ve seen some of my AI presentations I show a low-latency NSADI link between NonStop and the HPE Apollo HPC systems.  The additions of a low latency link between NonStop can provide a number of interesting capabilities to combine transaction processing with Machine learning models for fraud analysis, consumer history and being able to apply a next-best-action with customers.  In terms of healthcare an ability to provide a complete view of a patient with access to all articles and recommendations, based on diagnosed illness.  So a few of the many exciting possibilities.  It just so happens I was chatting with my good friend and colleague, Distinguished Technologist, Keith Moore about HPE’s recent Exascale announcement.  (see: https://bit.ly/3kbiMTy ). This comes from the acquisition of Cray.  As Keith pointed out, “Exascale is not just marketing hype. It’s an actual thing. This announcement is a truly unique new shift in computing and most especially for HPC.

Here is the technical breakdown in 4 points:

  1. A leapfrog performance JUMP.   This is not just a just a “chip-spin” level of performance change. It is 1000 times the existing limits!  The prior Cray system is 1015 FLOPS(petascale).  This announcement makes it 1018 FLOPS (extascale)!   In other words, 3 orders of magnitude performance increase! 3 ORDERS, yes, 1000 times faster than the last generation of Cray system!
  2. Parallelism/Messaging throughput (≈ 25+ TB/s!).
  3. Energy usage drops ≈ 95%.
  4. Memory-driven compute. “

The percentage comparisons above are between this new HPC and the current (last gen) Cray which was already a market leader.   Unlike normal product announcements, which are generally accretive, this one is actually a real tectonic shift in HPC.  I suspect it’s why HPE acquired Cray.

Why should the NonStop community care? NonStop and Cray are organized into the same product group at HPE. HPE NonStop can offer a RDBMS for use by HPC systems. A lot of HPC is data driven; albeit usually in memory and discovered/generated.   So relational and key-value databases are valuable as off-platform HPC tools. Also, very little HPC business usage is transactional. However, using a transactional engine, close-coupled to a HPC research machine, could be useful. A simple RDBMS attached to such a device might allow for hybrid business cases. I think this is particularly likely to be interesting for healthcare scenarios where some portions of the data discovery would be placed or pulled from a RDBMS.

This current announcement is mostly about Exascale and updates to the Apollo line. Obviously, these systems handle compute better than NonStop, but NonStop handles data better than these Exascale computers do.

The design of the initial Tandem/16 hardware was completed in 1975 and the first system was shipped to Citibank in May 1976. Tandem, the company, enjoyed uninterrupted exponential growth up through 1983 when Inc. magazine ranked Tandem as the fastest-growing public company in America.  Which company was number 2?  Cray.  How about that?

NonStop and Cray potentially better together; and perhaps, about time.

Print Friendly, PDF & Email
About Justin Simonds 12 Articles

Justin Simonds is a Master Technologist for the Americans Enterprise Solutions and Architecture group (ESA) under the mission- critical division of Hewlett Packard Enterprise. His focus is on emerging technologies, business intelligence for major accounts and strategic business development. He has worked on Internet of Things (IoT) initiatives and integration architectures for improving the reliability of IoT offerings. He has been involved in the AI/ML HPE initiatives around financial services and fraud analysis and was an early member of the Blockchain/MC-DLT strategy. He has written articles and whitepapers for internal publication on TCO/ROI, availability, business intelligence, Internet of Things, Blockchain and Converged Infrastructure. He has been published in Connect/Converge and Connection magazine. He is a featured speaker at HPE’s Technology Forum and at HPE’s Aspire and Bootcamp conferences and at industry conferences such as the XLDB Conference at Stanford, IIBA, ISACA and the Metropolitan Solutions Conference.

Be the first to comment

Leave a Reply

Your email address will not be published.


*