Visit additional Tabor Communication Publications
November 11, 2011
SC11, the world’s greatest yearly supercomputing show rolls into Seattle this week. To help prepare you for the big week, we have put together a list of the top 10 myths of the phenomenon that is SC. With a little discussion of each, hopefully we will bring out the good, the bad, the hopes and realities of SC. And, maybe, along the way we’ll see why SC matters so much to our community.
Myth 1: It’s all partying
Yes, it’s true, there are lots of networking receptions, meetings-over-a-drink, and business dinners held during the week of SC. Perhaps even one or two events that struggle to be described as anything other than a party. But hold on before you dismiss SC as merely a week of parties with a conference somewhere in the background. There will be several thousand attendees, over a hundred technical sessions, around 300 exhibition booths, and hundreds of hours of meetings held in the surrounding hotels and venues.
Before the exhibition has its grand opening on Monday evening, hundreds of staff from exhibiting organizations will have been working hard over the weekend to setup booths and demos. Lots of people will also have been attending tutorials, workshops, vendor meetings, and private business meetings over the weekend. So, for many attending SC11, Monday night is halfway into the week – not the first day – and a desire to wind down with a drink and friends is only fair.
Likewise, each day, people will have had to endure hours standing on booths sustaining extreme enthusiasm for their products, or sitting through slide decks that oscillate between no useful detail and far too much detail, or searching the exhibition or technical program for a solution to a problem they’re not sure they have yet. A sponsored social gathering is a good way to relax after all that - or to engage with more contacts if you are still feeling workaholic.
Myth 2: HPC center directors wander the exhibition floor with $100m to spend
Although every salesperson attending SC dreams of the big opportunity arriving from a casual encounter on the booth, I would bet that most people with $100m to spend have booked up a series of meetings with potential suppliers. After all, if you had given somebody that much money to spend, you’d hope they had a better procurement plan than to simply wander about the exhibit hall looking for a shiny toy that caught their attention?
Of course, in most cases, the supercomputer center directors don’t have the money to spend – government funding agencies have money to spend, supercomputer centers just help them spend it! Some government funding agencies use specialist advisors (like NAG or other independent consultants) to scout the market at SC to inform and plan for big procurements. In most cases, a competitive procurement process will actually spend the money – although meetings with vendors at SC are often used to support that process. But it is very rare that the $100m check is signed or announced at SC itself.
Myth 3: Industrial users of HPC have money to spend on every latest software gadget
It’s a popular hope of anyone selling products and solutions – especially in the niche software space. Especially those who think their niche is the answer to a perceived industry wide problem. Or those in the academic sector, who, against the challenges of sustaining software in the academic funding environment, turn in hope to industry as a promised land of plenty. Industry has untapped reserves of money. Or not.
Often users of HPC in industry are focused in small groups within companies and have operating budgets that would make national labs and academics cringe. Investing in software in industry can be a cultural and funding “adventure” too. The difference (to paint the extremes of the picture) is that the academic reward and funding system is determined to invest in new methods and implementations wherever possible (“research”) and avoid re-use or ongoing evolution, whereas the common industrial environment is determined to avoid investing in new software or major re-developments unless re-use or evolution is overpoweringly ineffective.
I have made these extremes with artistic license, but they serve the point – balance in speculation and stability is required – and software strategy is usually key to that balance in HPC.
Myth 4: Everyone knows everyone
Don’t be silly. It’s often said – I say it too – “supercomputing is a small community”. The extension, “everyone knows everyone” is stretching the credibility though, with several thousand attendees at SC11. But, there is a large degree of interconnection in the supercomputing community and so the community often feels close and small. Within the overall HPC world, there are a significant number of people who are very active and meet many times a year between SCs.
However, focusing only on those high profile members of the HPC world would be detrimental to the overall SC11 audience – many of whom will be new to HPC or on the peripherals of the traditional HPC community. SC must – and does – support all parts of the potential HPC community – new entrants, established centers, high profile researchers, and observers considering how HPC can help them. SC11 is a very social and friendly event for those who are seasoned HPC professionals – but we must make sure those who have poked a tentative toe in the water feel welcome and not overwhelmed.
Myth 5: SC week is the best time to issue a press release
Yes, it is. In the same sense that the best place to exhibit a tree is in the forest. You’d better have an impressive tree compared to the others - otherwise it’s just a tree with all detail forgotten. Each year, SC brings forth a flood of press releases, case studies, and so on. The only advantage to issuing your press release during SC itself is that, briefly, some of the core IT and even mainstream media spare some attention for supercomputing, so if you hit lucky you get a wider audience than normally.
But most press releases will simply be another small bubble in the seething froth of news frenzy that surrounds SC11 week. Plus, so many of your potential audience will be too busy creating their own news or standing on booths or in meetings to read anyone else’s news. Unless it’s big news of course. SC is a great time to spread and collect rumors though. Much more fun than factual press releases :-)
Myth 6: You can sustain 30+ meetings in a week and still be paying attention to that roadmap detail presentation on the last day
Do I really need to explain that one? See my SC11 diary series for what goes into some people’s SC11 planning. By Thursday afternoon’s presentation, vendor A’s plans to bring out widget B in 2019 with a feature of C=1.00234 rather than someone else’s 1.00235 is hard to focus on, compared to the impending flight home.
Myth 7: The technical program is the most important part of SC11
Not if you have $100m to spend on a supercomputer! Best to walk the show floor :-) or book some meetings with potential suppliers and your favorite HPC consultants. Joking aside, the technical program is actually very important. The competition for papers at SC is intense, which is promising for quality, plus there is a wide range of tutorials and workshops. Thus, there are excellent opportunities to catch up on the latest advances in HPC, learn new skills or disseminate your work.
But even the most academically focused attendee should never ignore the other two key parts of SC – the exhibition and the social/networking events. The exhibition can often provide as many gems of innovation as the technical program. Collaborations are sparked by conversations on the exhibition floor or at the social events.
Myth 8: The exhibition is the most important part of SC11
The exhibition can be a hive of activity – at its peak (Monday night’s grand opening) easily the most vibrant single event of the show, in my opinion beating even the technical program keynotes. The exhibition can also feel like a ghost town at times, with seemingly the only people in the vast hall those who have drawn booth duty.
Most of the time it is somewhere between the two, a steady stream of punters drifting or hurrying along the carpets, with eager booth personnel trying to woo them with logo’d giveaways, sweet smiles, or entrapping questions. But, as above, the exhibition is only one of the three important parts of SC - and the technical program and social events are just as important.
Myth 9: The whole HPC world will be at SC
Much as the organizers would like everyone associated with buying, using, selling or researching HPC products, technologies and services to attend, that is not the reality. As I wrote on my blog recently, there are many people for whom HPC is an essential part of their work or toolset, but is not an end in itself and thus SC is not a core event for them. Even within the community of HPC practitioners or researchers, many cannot attend SC every year.
But, against that, a huge proportion of the active HPC world will be at SC11 – most organizations providing HPC technologies, software, services, or research will be represented at SC11. So it will feel like the whole HPC world is in Seattle. And that’s the core attraction of SC – so many collaborators, prospects, potential solution providers, and contacts will there that it is hard to justify staying away.
Myth 10: The best SC11 myth of all …
To be told in person only. Find me in Seattle and I’ll tell you!
Jun 19, 2013 |
Supercomputer architectures have evolved considerably over the last 20 years, particularly in the number of processors that are linked together. One aspect of HPC architecture that hasn't changed is the MPI programming model.
Jun 18, 2013 |
The world's largest supercomputers, like Tianhe-2, are great at traditional, compute-intensive HPC workloads, such as simulating atomic decay or modeling tornados. But data-intensive applications--such as mining big data sets for connections--is a different sort of workload, and runs best on a different sort of computer.
Jun 18, 2013 |
Researchers are finding innovative uses for Gordon, the 285 teraflop supercomputer housed at the San Diego Supercomputer Center (SDSC) that has a unique Flash-based storage system. Since going online, researchers have put the incredibly fast I/O to use on a wide variety of workloads, ranging from chemistry to political science.
Jun 17, 2013 |
The advent of low-power mobile processors and cloud delivery models is changing the economics of computing. But just as an economy car is good at different things than a full size truck, an HPC workload still has certain computing demands that neither the fastest smartphone nor the most elastic cloud cluster can fulfill.
Jun 14, 2013 |
For all the progress we've made in IT over the last 50 years, there's one area of life that has steadfastly eluded the grasp of computers: understanding human language. Now, researchers at the Texas Advanced Computing Center (TACC) are utilizing a Hadoop cluster on its Longhorn supercomputer to move the state of the art of language processing a little bit further.
05/10/2013 | Cleversafe, Cray, DDN, NetApp, & Panasas | From Wall Street to Hollywood, drug discovery to homeland security, companies and organizations of all sizes and stripes are coming face to face with the challenges – and opportunities – afforded by Big Data. Before anyone can utilize these extraordinary data repositories, however, they must first harness and manage their data stores, and do so utilizing technologies that underscore affordability, security, and scalability.
04/15/2013 | Bull | “50% of HPC users say their largest jobs scale to 120 cores or less.” How about yours? Are your codes ready to take advantage of today’s and tomorrow’s ultra-parallel HPC systems? Download this White Paper by Analysts Intersect360 Research to see what Bull and Intel’s Center for Excellence in Parallel Programming can do for your codes.
Join HPCwire Editor Nicole Hemsoth and Dr. David Bader from Georgia Tech as they take center stage on opening night at Atlanta's first Big Data Kick Off Week, filmed in front of a live audience. Nicole and David look at the evolution of HPC, today's big data challenges, discuss real world solutions, and reveal their predictions. Exactly what does the future holds for HPC?
Join our webinar to learn how IT managers can migrate to a more resilient, flexible and scalable solution that grows with the data center. Mellanox VMS is future-proof, efficient and brings significant CAPEX and OPEX savings. The VMS is available today.