Q&A with John Shalf, chair of the Disruptive Technology showcase at SC10

Print Friendly, PDF & Email

I’ve mentioned the Disruptive Technologies event at SC10 a few times recently, and I thought it might be helpful for you guys and gals if we dug in and explored the event, its background, and what it’s all about in a little more depth. SC10 and John Shalf, from Lawrence Berkeley National Laboratory and the chair for Disruptive Technologies at this year’s show, sat down with us over email to talk with insideHPC about the plans for this year’s event.


insideHPC: What is a disruptive technology in the context of this event at SC?

SC10 logo

John Shalf: “Disruptive technology” refers to drastic innovations in current practices such that they have the potential to completely transform the high-performance computing field as it currently exists — ultimately overtaking the incumbent technologies or software tools in the marketplace. Disruptive Technologies, which has taken place as part of SC since 2006, examines new computing architectures and interfaces that will significantly impact the high-performance computing field throughout the next five to 15 years, but have not yet emerged in current systems.

insideHPC: Why is it relevant to the SC conference series?

Shalf: We examine Disruptive Technologies to bring awareness of those ideas that may be coming down the road so that feedback can be provided at critical stages resulting in the idea’s maximum value when the technology is adopted. An example of a disruptive technology is commodity cluster computing. Over time, it has transformed the hardware and software ecosystem for high performance computing.

The focus of the SC conference can often skew towards near-term considerations. However, the most exciting research ideas and boldest departure from business as usual comes from groups that are looking at ideas that are at least a decade out. Disruptive Technologies creates a venue to bring early stage technologies out into the open to challenge our pre-conceived notions on how things are done, to give us the breath of options and possibilities and foster discussion about the future of computing.

insideHPC: Can you tell us a little about the history of Disruptive Technologies as a formal part of the SC program?

Shalf: The Disruptive Technologies exhibit originated within the SC06 conference “Exotic Technologies” thrust area. A technology may start out as exotic, niche or situational, but when it provides a new and alternative option for HPC that others need to acknowledge then we have a disruption.

The exotic technologies exhibit was crafted as a venue to discuss ideas and technologies that are 10-15 years out into the future – too far out to be in a product. This is indeed where some of the most interesting ideas and boldest thinking in HPC technology are taking place. Many of the exhibits on the SC show floor have increasingly focused on the near term issues and existing product roadmaps. However, we know that the pathway for going from research to a product can be very long. Some of the most interesting ideas and technologies in development are not even on the product roadmap. In 2007, the exhibit’s name changed to “Disruptive Technologies” in homage to Clayton Christensen’s 1997 book, The Innovator’s Dilemma, where the term was first coined. The exhibit today continues to serve as a forum for examining technologies that may significantly reshape the world of high performance computing.

insideHPC: Is there a particular focus or theme that you are interested in showcasing this year?

Shalf: This year we want to highlight technologies, which can be hardware, software, communications, power or thermal management, that will be enabling exascale computing. By all accounts, the path to exascale computing will require many highly disruptive technology phase transitions. Yet there are many enormous hurdles that remain to be solved over the next decade to overcome power, performance, and costs to get to a practical exascale computing platform by 2018. Therefore, any technology that is able to overcome these hurdles will be “disruptive” by definition.

insideHPC: What technologies have been showcased in the past that were especially interesting or, with the benefit of hindsight, especially prescient?

Shalf: The early discussions of 3D chip stacking, low power non-volatile memory, and silicon photonics technology were particularly interesting in retrospect. At the time, they seemed like exotic packaging technologies, and their importance was not broadly recognized at the time. However, after 3 years of deep investigation of hardware constraints, and a more complete understanding of the technology options required to get to exascale, these technologies have emerged as being on the critical path to overcome many of the challenges for exascale computing.

insideHPC: Is there typically good international participation?

Shalf: We are very much looking to increase international participation. Just as the SC show has grown in size and in its international scope, we hope that the Disruptive Technologies program will also attract greater international attention.

insideHPC: When is the showcase open? Where can attendees find it?

Shalf: The Disruptive Technologies exhibit will have a highly visible location on the exhibit floor, so will consequently open and close with the SC10 exhibits. In addition, we will have a panel on Friday on the DARPA Ubiquitous High Performance Computing (UHPC) program, which is designed to foster new innovative projects to develop radically new computer systems that overcome the challenges of efficiency, dependability, and programmability anticipated in the exascale era. The panelists, who lead the selected UHPC teams, will discuss the perspective on how to address these challenges and their comprehensive hardware/software strategy they developed for the UHPC program.

insideHPC: How can interested folks be a part of this discussion?

Shalf: Companies or organizations that wish to participate in the Disruptive Technologies exhibit should apply at the submission website https://submissions.supercomputing.org/, by August 5 2010.

The submission form is very straightforward. We just need a title, contact information, 250 word abstract describing your technology and why it is disruptive, and 50 words to describe what you need to show off your invention (e.g. power hookups, projectors, how much table space in our booth). Finally, you can upload any supporting documentation describing your invention (extra information for the committee to evaluate the disruptiveness of your technology).

Anyone considering applying but who has questions can get in touch with us at disruptive-techs@info.supercomputing.org, and we will be happy to answer any questions about the exhibit or your submissions.