ISC 2017, SUSE, and the HPC Tesseract

The ISC High Performance conference – one of the most forward-thinking gatherings of supercomputing aficionados in the world – has just gotten underway in Frankfurt, Germany.  With many supercomputing facilities being represented here, and the majority of those employing SUSE Linux Enterprise as their operating system of choice to power their HPC software infrastructure, this is a significant opportunity to discuss leading-edge HPC projects with key players and partners – strategically important for the advancement of science and technology.  This year’s pervasive area of research opportunity focuses on artificial intelligence and machine learning.

Transformation and the HPC Tesseract

I refer to it as the “HPC tesseract” to emphasize the expanding importance of deep learning across many industries today – including healthcare, banking and finance, manufacturing, pharmaceuticals, and more.  The tesseract, also referred to as a hypercube, illustrates a four-dimensional geometric view.  When describing HPC, the 3D cube encompasses the hardware and software symbiosis required for advanced clustering and multi-processor support required to handle today’s massively complex computational simulations.  The HPC tesseract’s fourth dimension metaphorically refers to the growth of AI on a time scale defined by machine learning, deep learning, and expansion of high performance data analytics.

Our theme at this year’s conference is around transformation and adoption of HPC.  SUSE continues to transform as well within the HPC space – from managing departmental and enterprise clusters to enabling HPC as a Service to transforming to support new HPC architectures.  Coinciding with ISC this year is the worldwide release of the latest “Transformers” movie.  We’ve adopted the Transformers within our SUSE booth not only to represent the transformation that SUSE has seen and enabled for HPC adoption, but also to signify the growing interest in AI that is becoming so much more pervasive across industry projects.

The adoption of high performance data analytics, big data and a more cohesive open source HPC stack has fundamentally changed the landscape for building AI and deep machine learning applications.  Along with a strong HPC operating system and infrastructure coupled with advancements in AI, machines are becoming more intelligent at a faster pace.  They are improving their ability to recognize patterns, iteratively learn from data analysis and sharpen predictive analytics.  Industries starting to benefit from these AI applications include banking and finance, pharmaceuticals, shipping and transportation, healthcare, manufacturing and agriculture. With SUSE Linux Enterprise Server for High Performance Computing, we now have greater opportunities to leverage underlying hardware and power the new wave of AI and data-intensive applications across industries.

The Future of HPC Is Here

High-performance computing (HPC) has become a new reality. Predictions are coming from all corners of the industry, but the one idea that remains consistent is that HPC is more accessible and mainstream than ever before. Not long ago, HPC technology was limited to the world of large research institutions, governments, and massive enterprises. Today, HPC is finding practical applications in a variety of sectors including financial services, media and entertainment, healthcare and pharmaceuticals, and weather and climate forecasting.

The research supports this trend. Market research firm Hyperion (formerly IDC HPC analyst team) says that HPC has been one of the world’s fastest-growing IT markets. Worldwide revenue between 2001 and 2016 more than doubled from $4.8 billion to $11.2 billion with a predicted value of $14.8 billion by 2021. Add software, storage, and services and the 2021 forecast approaches $30 billion.[1]

In its list of top 10 predictions for 2017, Gartner names artificial intelligence and machine learning, both powered by HPC. Gartner contributor Kasey Panetta writes, “AI and machine learning have reached a critical tipping point and will increasingly augment and extend virtually every technology enabled service, thing, or application. Creating intelligent systems that learn, adapt, and potentially act autonomously rather than simply execute predefined instructions is primary battleground for technology vendors through at least 2020.”[2]

Disruptive Technologies Fuel HPC Growth

This idea of intelligent systems that learn and act autonomously is apparent in several disruptive technologies made possible by HPC. With HPC more accessible and mainstream, we’re seeing great innovation in cognitive computing, advanced robotics, and autonomous transportation.

Cognitive computing and intelligent software systems have many practical applications. Created to perform knowledge work, these systems were built using HPC technologies such as artificial intelligence, machine learning, high performance data analytics for big data, and natural language interfaces.

In healthcare, these systems aid in diagnostics and drug research, potentially increasing quality of life for millions of patients. Other applications include smarter learning in education, intelligent investments in finance, and assisted contracts and patents in the legal field.

Supercomputing with SUSE for HPC

While the HPC market is growing, many would contend that it is still in its infancy. As a result, the HPC market faces many challenges. HPC stack components are provided by multiple, various organizations representing a mix of large and small vendors. And applications must be updated to leverage an HPC infrastructure if they are to bring value to HPC technologies that have been historically segmented into commercial and scientific applications.

With its industry-leading Linux operating system for the enterprise, SUSE is working to resolve these challenges and drive growth in HPC and disruptive technologies like artificial intelligence and machine learning. SUSE Linux Enterprise Server for High Performance Computing is a specialized version of SUSE Linux Enterprise Server that focuses on the challenges of transaction-intensive computing environments. With selected tools and components used in HPC environments, SUSE for HPC offers support for the latest hardware infrastructures for high-performance computing.

The SUSE vision for HPC includes increasing market awareness and positive perception of SUSE in the HPC community. With strong credibility and references for high performance data analysis workloads, SUSE offers leadership training in traditional HPC deployments, ARM HPC deployments, and differentiated HPC offerings across multiple industries and technologies.

With Linux being the leading operating system for HPC-enabled disruptive technologies, SUSE works to accelerate innovation with a reduction in administration. This allows users to realize faster time to value through simplified configuration of HPC workloads. The solution enables users to innovate new HPC technologies by:

• Simplifying administration with a highly scalable, high-performance, open source operating system that harnesses the power of the supercomputer for day-to-day business
• Enabling faster time to value with simplified configuration, easier access, and reduced maintenance
• Accelerating innovation with a strong ecosystem of partners that transforms the enterprise storage infrastructure to reduce costs while providing unlimited scalability
• Reducing the time and effort required to set up and maintain HPC clusters
• Ensuring that all components of the HPC stack work together

Conclusion

The outlook for High Performance Computing is strong and optimistic, with uses cases like AI and Machine Learning garnering more and more attention, and wider HPC adoption across many industries already utilizing data analysis today.  With open source communities like OpenHPC, SUSE’s growing HPC partner ecosystem and a cohesive HPC stack – the future is bright.  With many real-world applications that bring HPC beyond the realms of government and academia, HPC has become a new reality with great potential for growth. As HPC technologies become more prevalent within commercial enterprise data centers, HPC has the potential to transform big data analytics and generate disruptive technologies that will become a new way of life.

For additional product information, please go to the SUSE HPC Solutions page at https://www.suse.com/solutions/sle/high-performance-computing/ , the SUSE HPC product page at https://www.suse.com/products/server/hpc/ , or contact me at Jeff.Reser@suse.com .

@JeffReserNC

[1] Hyperion Research, “Linux and Open Source Are Driving HPC into New High-Growth Markets,” April 2017.

[2] Kasey Panetta, Gartner, “Gartner’s Top 10 Strategic Technology Trends for 2017,” October 18, 2016.