print 'n fly - sc13

Print ‘n FlyExclusive Interview

The Fast Data ImperativeMichael Kagan,

CTO at Mellanox

Interview with Bill Gropp

PlusThe Complete History

of Keynotes

ANNIVERSARY25TH

Brian Sparks and Scot Schultz on

The Secrets of Technical Marketing

Sci-Fi Original: The Observer

EffectBy Rich Brueckner

Restaurants, Entertainment, and More

2Print ‘n Fly | November 2013

3 The Fast Data Imperative 5 Interview: Bill Gropp on the 25th Anniversary of SC 7 Technical Marketing in the Age of HPC 9 Local’s Guide to Restaurants in Denver 10 Move Over, HPCers: Another Wave of Immigrants is Hitting Your Shores 11 SC 25th Anniversary – The Complete History of Keynotes 12 Local’s Guide to Bars and Entertainment in Denver 13 Brian Sparks and Scot Schultz on the Secrets of Technical Marketing 16 An Update to the Exascale Progress Meter 19 SC13 Comes to Denver, Colorado 21 Sci-Fi Original: The Observer Effect

Mellanox Welcomes You to Denver and SC13

Come See Us at Booth #2722 www.mellanox.com/sc13

Contents• • • •

• •

•


insideHPC: With the 25th anniver-sary of the SC conference this year, could you please share your perspec-tive on how far we’ve come in system interconnects over the past quarter- century?

Michael Kagan: As technology has evolved, the interconnect has played a paramount role to bringing new concepts and enabling new lev-els of system performance, efficiency and scalability. We have witnessed the migration of HPC interconnect tech-nology intoWeb 2.0, cloud, and data intelligence systems to handle the explosive growth of data. We are in the era where the interconnect is critical for so many applications. For instance, real-time data analytics of unstructured data on a global scale was only a concept 25 years ago, however today, real-time access to data is literally a world econom-ic dependency. When you think about it in these terms, the interconnect is just as important to our global economy as water is to the fish in the sea.

insideHPC: In terms of today, how well-entrenched is the lat-est generation of FDR InfiniBand out there in the TOP500 supercomputers?

Michael Kagan: Mellanox FDR InfiniBand systems grew over 3X from June’12 to June’13. We are pleased with the adoption of our technology and believe it will continue to be the performance leader in the TOP500. As Petascale ca-pable computing is becoming more common-place in the TOP500, InfiniBand is also the leading interconnect with 16 out of the 33 systems in the June 2013 list, and what’s more, FDR InfiniBand connects the four fastest InfiniBand systems on the list. From our perspective, we believe the TOP500 is a good metric to show that the HPC commu-nity has a high regard for industry-standard, high perfor-mance interconnect solutions that are well supported by an ecosystem of open-source.

insideHPC: What comes after FDR? Will your ConnectX Ethernet likely take a leap at that point as well?

Michael Kagan: The next generation 100Gb/s technol-ogy comes from the architectural building blocks of ConnectX-3 VPI and Connect-IB FDR technology, and is planned to be released in the 2014-2015 time frame. Both, our InfiniBand and Ethernet products, deliver the best de-liver the best price/performance in the marketand we see increased adoption of both. In fact, Mellanox sold more

than one million end-points last year; around 10% of the server market.

insideHPC: The SC13 conference is loaded with sessions on Exascale. Will InfiniBand have a role to play as we get to 10^18th flops sometime in the next 7–10 years?

Michael Kagan: We believe we are point-focused on Exascale and are on target to address the challenges it pres-ents. This includes advanced data protection mechanisms, support across the various network topologies, instruction set architectures and acceleration technologies, improving increased bandwidth and providing the lowest possible latency, but particular attention on improving perfor-mance at scale will be a key issue. We participate in sev-eral of the Exascale programs and plan to continue and in-vest in delivering faster and faster interconnect solutions to pave the way for Exascale.

insideHPC: Mellanox has made some Photonics acquisitions recently. Is this the future and are we reaching the limits of silicon?

Michael Kagan: Yes, Mellanox has made two significant acquisitions in the area of silicon photonics and VCSEL based technologies. We acquired a company called Ko-tura, which brings a world-class team of silicon photonics expertise. We also recently acquired IPtronics, a leading fabless supplier of low-power, high-speed analog semi-conductors for parallel optical interconnect solutions. We believe these investments are strategic to enable 100Gbs data rates and beyond.

ANNIVERSARY25TH

My fondest memory of SC? My first SC04 in 2004 in Pitts-burgh was the most impressive event that I attended as a young student at that time. The gigantic machinery that was brought just to showcase supercomputing demon-strated how important and big the HPC field really was. For me, this experience was a game, or better, career-changer. I decided to focus my research on HPC because this is a field where one can have real impact. Since then, I attended every SC!

– Torsten Hoefler, ETH Zürich

What is your fondest memory of the SC conferences?

The Fast Data ImperativeAn Interview with Mellanox CTO, Michael Kagan

• • • •

• •

•

Michael Kagan


insideHPC: What happens to Mellanox if x86 vendors start building InfiniBand interfaces directly onto their processors?

Michael Kagan: It is all about bringing the best solution to the market. To date, Mellanox has delivered 7 generations of advanced interconnect solutions and is a generation or more ahead of the competition. We believe that we will maintain and even extend the gap into the future. Only Mellanox provides a complete end-to-end solution for 40 and 56Gb/s, and it is not an easy task. Furthermore, once you add the software complexity into the mix, you will see that it is not easy for other companies to bring the solu-tions that Mellanox has today.

insideHPC: As CTO, what is your toughest challenge in pushing the limits of communications technology? (Resilien-cy, component cost, bandwidth, power consumption?)

Michael Kagan: The toughest challenges are usually only uncovered when you actually try pushing the limits of technology, which coincidentally is a driving force at Mellanox. By far, what is the most challenging is to truly understand the problem for which you are trying to pro-vide a solution. There are many adjacent issues that need to be considered; such as security in the cloud, scalability to hundreds of thousands of end-points and power effi-ciency to name a few.

ANNIVERSARY25TH

My fondest memory of SC? I first attended SC in 1989, to present a technical paper. Back then we used 35mm slides (instead of Powerpoint), and since my paper was about de-bugger visualization techniques the illustrations were criti-cal. About halfway through, the projector lamp burned out. There I was, using my hands to “draw” figures in the air in order to explain what the audience should have been see-ing on the screen. I was young enough to be embarrassed, until I realized that the audience actually became more re-ceptive because they sympathized. That single presenta-tion led to a number of collaborations and friendships. I’ve never missed an SC since.

– Cherri Pancake, Oregon State University



SC13 marks the 25th anniversary of the supercomputing conference series. To get a perspective on the history of the show, we caught up with Bill Gropp, the SC13 Confer-ence Chair.

insideHPC: When you took over as conference chair this year, what were the top things you set out to accomplish? How did it go?

Bill Gropp: SC is a unique — and uniquely valuable — event in our community, with a long history of making a difference. When I accepted the challenge of chairing the conference, I set the bar high: I wanted to make SC even better than it is already. Some of my most important goals were:

• Increase the recognition of the widespread contribu-tion of HPC to all aspects of society

• Make SC the most desirable place to present and pub-lish the best technical work in HPC.

• Make SC an even more effective place for everyone involved in HPC to collaborate and make new con-nections.

So far, I think that it is going well, but you and your readers will have to tell me how well my committee and I have suc-ceeded in making SC even better. And I believe that we’ve made the planning process better; for example, we’ve re-duced the number of planning meetings and we’re using collaboration tools to better organize and prepare.

insideHPC: What new things can attendees expect to see this year at SC13?

Bill Gropp: The model that has evolved for the conference is that we have a core of excellent program elements — tutorials, technical papers, the exhibition, and so on — and then we innovate around that core. Each year the commit-tee tries a few new things and, while some don’t work out, those that do work out may ultimately be added to the core of the conference. This helps us ensure that each SC is familiar, but stays fresh and exciting.

This year we’ve added some truly interesting aspects to the program:

• There are more ways to see HPC in action with Emerging technologies and the HPC Impact Show-case, both on the exhibit floor. Emerging technologies is a peer-reviewed addition to the technical program that showcases new developments, permitting the live demonstration of new technologies and ideas. The HPC Impact Showcase highlights the use of HPC, from consumer goods to industrial applications. It provides a venue for the users of HPC to interact with the developers and providers of HPC technol-ogy, and will include many large corporations using HPC that I think attendees will find very interesting.

• Looking back over the past 25 years, it is clear that SC has been the launching point for many of the techni-cal innovations that have reshaped HPC. To recog-nize this history of impact we have started “Test of Time” award, which will be presented for the first time this year. This new award will be presented each year to a paper presented at SC that has had the greatest sustained impact on our field.

• HPC Interconnections is our new name for the SC Communities program, to emphasize the focus on providing an “on ramp” into the excitement of our field. This year we are integrating it even more tight-ly with the technical program. I encourage everyone to have a look at the offerings of the HPC Educator and Broader Engagement programs; the sessions be-ginning Tuesday are open to all Technical Program attendees.

• In celebration of the 25th anniversary of SC we have special exhibits, including some at participating ex-hibitors, and a panel discussing the past and future of our field.

• This year we will have four major awards present-ed at SC — in addition to the Cray, Kennedy, and Fernbach awards, Prof. Kathy Yelick will receive the ACM-W Athena Lecturer Award. The Athena award is presented annually to an outstanding female re-searcher who has made fundamental contributions to computer science. This is the first year a researcher from our community has been selected for an Athena, but I hope this is the start of a regular feature in our conference program.

Interview: Bill Gropp on the 25th Anniversary of SC• • • •

• •

•

Bill Gropp

CONFERENCE DATES November 17–22, 2013

EXHIBITION DATES November 18–21, 2013

Looking back over the past 25 years, it is clear that SC has been the launch-ing point for many of the technical innovations that have reshaped HPC.

“ “


insideHPC: What is your fondest memory of the SC conferences?

Bill Gropp: There are so many, it’s hard to pick just one. Here are few that come to mind:

• MPI really started at an ad hoc meeting at SC in Min-neapolis in 1992. It was at this meeting that I made the commitment to begin an implementation of MPI as the MPI Forum, as the group became known. That implementation became MPICH. This is also an ex-ample of the positive impacts of major meetings like SC — it provides an effective venue for groups to gather and accomplish something that none of them can do on their own.

• As a member of a panel on multicore organized by Thomas Sterling and including Ken Kennedy; the room was packed and the discussion lively. This was the last time I saw Ken; he was both a great research-er and a true gentleman. I’m glad that his memory is honored with the ACM/IEEE-CS Ken Kennedy award.

• Clayton Christiansen’s keynote in 2010. The material was fascinating and his personal story was compel-ling.

• The SC Committee family. Few of the attendees see the dedication of the many volunteers that makes SC possible. Becoming part of this group, and being honored with the responsibility of SC13, is something I will always treasure.

• I’ve had fun wandering the exhibit floor looking for new parallel computer vendors and asking about their MPI implementation. The fun was how long it took until they admitted it was MPICH.

insideHPC: How do you think the conference has changed over the years?

Bill Gropp: One obvious change is the size — SC is much larger than it was when I first attended, in 1992. The tech-nical program has become very strong. The focus of the technical program has also changed, with less focus on compilation and languages and more on massive paral-lelism. Another change is in the composition of the exhibit floor, especially from the research exhibits, which have become more active and dynamic, and I believe have cre-ated an atmosphere that encourages more collaboration, exploration, and friendly competition.

insideHPC: How important is the SC conference to the global HPC community?

Bill Gropp: SC is essential. An important ingredient in scientific progress is the efficient exchange of ideas, and the best way to get that free thinking, efficient exchange is to immerse the best researchers, users, developers, and teachers in a stimulating sea of ideas — whether it is pre-sentation of new results in papers or posters or in the dem-

onstration of new ideas on the exhibit floor — and give them the opportunity to interact. Big meetings such as SC are one of the few places that you can bring all of these elements together.

insideHPC: Which conference / year was your favorite - and why?

Bill Gropp: You like to ask the hard questions! The easy answer is 2013 in Denver — it’s been great to be the gen-eral chair and to see everything come together. Denver is a great place to hold a meeting like SC. But if we exclude the current year, I’ll pick 2009 in Portland. As chair of the technical program, I was deeply connected with the pro-cess of making the conference a success, and my technical program committee was fabulous. I also have a soft spot for Portland — I had family there and they have one of the world’s best bookstores — Powell’s.

insideHPC: What effect do you think the Government Shut-down will have on attendance at SC13?

Bill Gropp: That will depend on the next few weeks. We’ve already issued a statement on the impact of the shutdown on SC13, and right now I can say that we have not seen any significant impact. Registration is still very strong both for the technical program and for the exhibit floor. We’re also fortunate that SC starts in mid Novem-ber; there is plenty of time for the chaos created by the shutdown to settle. We also extended early registration for those affected by the shutdown, and we’re already seeing registrations from people who could not register during the shutdown.

insideHPC: At this point, has anyone already cancelled talks or exhibits as a result of the government shutdown?

Bill Gropp: Not that I’m aware of. For those that are di-rectly affected, we are waiving the cancellation fees for now so that no one will feel forced to cancel, and we also extended the early registration pricing. Indications are that those directly affected were patient and waited to see how and when the shutdown ended; we’ve already received thanks from some of those affected for the steps that we took to accommodate them.

An important ingredient in scientific progress is the efficient exchange of ideas, and the best way to get that freethinking, efficient exchange is to immerse the best researchers, users, developers, and teachers in a stimu-lating sea of ideas and give them the opportunity to interact.

“

“


SC13 will have over 360 exhibits this year. In such a vibrant and growing technology space, how does your company stand out with marketing programs that deliver results? To get a perspective on this, we caught up with a seasoned HPC marketing veteran, Mike Bernhardt. Many of you know Mike’s name as the Pub-lisher of The Exascale Report, but he is also a highly regarded strate-gic marketing and communications consultant and sits at the helm of Team Libra Communications.You might also remember Mike as the SC09 Communications Chair, a year where I think the conference really knocked it out of the park.

insideHPC: What would you tell these marketing guys out there who are fighting for internal resources?

Mike Bernhardt: Executives who may be one or two steps removed from marketing have to realize that within the HPC community, there is a certain level of awareness and brand recognition activity that’s expected from a vendor who wants to be taken seriously in this space. It’s a fast paced market in many ways. People are constantly work-ing on proposals, responding to RFPs, and working on a wide range of procurement deals. It’s actually very easy for an organization in acquisition mode, within the HPC community or at the fringe, to overlook one of the key HPC solution providers. And even if they’re reminded of a good vendor candidate along the way, it may be too late for an active RFP.

So staying in front of those folks on a continual basis is absolutely critical within the HPC segment. I think that some organizations tend to think of marketing as some-thing that is nice to have, and they overlook how critically important it is. They need to have their messaging, posi-tioning, and competitive posturing out in front of people all the time through multiple channels.

So let’s consider a customer organization out there that has a procurement team, and they’re evaluating alterna-tives for a new computing or storage system. If they’re considering the options, they would probably start with the vendors that they are most familiar with. But if these decision-makers are constantly seeing the messaging from one of the newer, more aggressive players, they are often inclined to take a look at that. And here is the key--they tend to do that process through peer-to-peer communica-tion. If their influence partners are not aware of the vendor

or solution provider, the credibility factor of that vendor or product immediately slips.

In the HPC Segment, decision makers don’t necessarily start by clicking an ad or reading a story, and then signing up for more information. But such impressions can inspire them to talk to their peers and colleagues to determine if that product is a viable solution for them. So from what I’ve seen over my 26 years in this community, this is very differ-ent from the way that a potential customer might respond to an ad in the mobile or enterprise computing spaces.

insideHPC: So you’re saying there’s a different pathway to sales here in HPC?

Mike Bernhardt: What I’m saying is that, as a vendor, un-less I’ve made a strong impression upon these folks to get to their top-of-mind, they won’t be seeking that counsel with their peers.

To make matters even more complicated, when you look at a market like HPC that has some long-standing vendors, those companies tend to become known for one thing. And that makes it especially tough for them when they add new capabilities and technology solutions outside of that known space. Their marketing teams have to really step up to overcome that notion that your company just does an updated version of what they did 20 years ago.

So even with the cornerstone organizations in HPC, they have to tell us what they’re about today. And they can’t depend on one editorial piece, trade show, white paper, or press release to get the job done. If they don’t hit multiple channels, it’s just way too easy for decision-makers to miss the message.

So beware the checkmark that says, “That’s a done deal. We communicated that.”

insideHPC: How has marketing in HPC changed over the years you’ve been serving the community?

Mike Bernhardt: When I first entered the HPC space do-ing marketing programs some 26 years ago, technical mar-keting was just starting to come into its own. At that time, I don’t think marketing people were all that welcome in a technical community that just wanted the facts.

Technical Marketing in the Age of HPCAn interview with Mike Bernhardt from The Exascale Report

• • • •

• •

•

Mike Bernhardt

ANNIVERSARY25TH

1992. Vikings and Cheerleaders.

– Mike Bernhardt

Which conference or year was your favorite, and why?


And so I think you had to pay great attention to balanc-ing your marketing programs, even though you wanted to beat your chest and brag. You had to learn to temper that hyperbole quite a bit and as a result, the tone got increas-ingly conservative in the early 1990s.

And then things started to loosen up a bit after that in the mid-90’s and we saw more flashy marketing approaches (think animatronic robots in the SC95 Cray booth for ex-ample) and people started adding more fun things into their marketing programs.

But in the early 2000s, I think that it was about the time when Grid was getting very popular and everything just got blown out of proportion. There was even a one-time Grid trade show in Philadelphia that was considered somewhat of a disaster that really drove the point home. You just couldn’t separate the facts from hype, and that left the technical computing community with a very bad taste in their mouths. When that happens, they tend to be less open to receiving almost any message from market-ing. They then go tighter into the huddle with their peer groups to talk through what’s going on with technology, while discounting almost everything that comes from a marketing organization.

So it goes in cycles. These days, I think we’re seeing things loosening up to the point where I think we can start hav-ing fun with marketing again. And there are a couple of things affecting that. The old guard of supercomputing is thinning out, and there’s a new generation of HPC pro-fessionals coming in. And these younger people (age, ex-perience, skills) are often more open to a more fun and lighthearted approach to things, as long as it’s all based on reality, facts, and real-world comparisons.

With that in mind, I think a company that learns to present the facts in a fun and entertaining way will see their mind-share increase out there. And I think we are going to see a change in the way companies are doing their marketing as a result. It’ a positive change, and I think its going to be good for the HPC community.

insideHPC: How much does marketing really help if an HPC vendor’s target prospects are within the national labs?

Mike Bernhardt: That’s something I’ve always been fas-cinated with. While many of my marketing peers feel you don’t need to market to the labs, I truly believe that mar-keting does matter there. But it’s an indirect influence that in many cases, really helps the lab procurement teams to confirm their decisions. In general, the folks from the labs are extremely well connected through a viral network that’s beyond the reach of many vendors. But, if they do need an update from a vendor, they just pick up the phone and the next day the vendor is out there to give a brief-ing. And I know you’ve seen that at insideHPC, where someone from the labs saw a new technology in a booth video and was on the line with the vendor that very day.

That just speaks to the tremendous power and influence you can have if you deliver the message to the customer in the right way. But let me tie back to why solid, credible marketing is important to the labs. And this is true on an international level.

The future for any of the research labs depends on their ability to procure new funding. And that is obviously based on both the real and the perceived success of the programs they launch. When they request millions of dol-lars for a program that is based on technology from a par-ticular company, they need to know that the companies and the products they intend to work with have the right appeal and the support of their community. So the way a vendor’s marketing programs are perceived, with items such as success stories, case studies, and customer-driven white papers really helps the labs in making their own case when they go out to look for their funding.

Mike Bernhardt has held staff positions with a number of lead-ing edge companies including Prime Computer, DEC, Multiflow Computer and Intel. As a consultant and agency owner, he has been responsible for branding, advertising, marketing and com-munications campaigns for dozens of HPC companies over the past 25 years.If you don’t know Bernhardt’s name, you will prob-ably recognize his face. He truly is one of the HPC veterans. This is Mike’s 26th consecutive year for attending the SC conference.

www.mellanox.com/connectibConnect with us @ MellanoxTech

100Gb/s interconnect throughputUnlimited scaling with new transport layer technology>137M messages per second

Connect-IB FDR InfiniBand Adapter

Server and Storage AdapterThe World’s Most Scalable


Larimer Square – Mecca For Foodies

Corridor 441433 Larimer Street Denver, CO 80202

(303) 893-0044

Cru Wine Bar1442 Larimer Street Denver, CO 80202

(303) 893-9463

Osteria Marco1453 Larimer Street Denver, CO 80202

(303) 534-5855

Row 14 Bistro & Wine Bar

891 14th Street, #100 Denver, CO 80202

(303) 825-0100

Worth a Cab Ride

Table 6 Local’s best-kept secret

609 Corona Street Denver, CO 80218

(303) 831-8800

Carmine’s on Penn Family Italian

92 S Pennsylvania Street Denver, CO 80209

(303) 777-6443

Fruition RestaurantLocal farm-grown goodness

1313 E 6th Avenue Denver, CO 80218

(303) 831-1962

Steak

The Capital Grille1450 Larimer Street Denver, CO 80202

(303) 539-2500

Seafood

Jax Fish House1539 17th Street

Denver, CO 80202

(303) 292-5767

Oceanaire Seafood Room

1400 Arapahoe Street Denver, CO 80202

(303) 991-2277

Best Place for a Group of 8 or More

Denver Chop House & Brewery

The Chop House has a private room you can reserve

for no charge

1735 19th Street Denver, CO 80202

(303) 296-0800

Local’s Guide to Restaurants in Denver• • • •

• •

•

Forget Yelp. We talked to the locals to get this listing of the best restaurants in Denver.


The global HPC market’s enormous expansion from about $2 billion in 1990 to $21.9 billion in 2012 has been heavily driven by successive waves of new users, with each new group of arrivals carrying its own expectations and opportunities. Growth-bringing newcomers have generally been welcomed to the fold by prior HPC users — especially those who remem-ber the post-Cold War market dip of the early 1990s, when there were serious concerns about the long-term viability of supercomputing. Xenophobia from the HPC establish-ment has been blessedly rare (although I recall giving a talk on SMB users at ISC a few years ago and hearing a few HPC stalwarts in the audience sniff, “That’s not high performance computing.”).

Let’s look at the waves of new users and their require-ments. In 1976, the CRAY-1 supercomputer was deliv-ered to its first customer, Los Alamos National Laboratory (LANL), as a blazingly fast hardware platform with no op-erating system. Not to worry: LANL and others in the first wave of HPC users, primarily government and university researchers, typically had enough in-house technical sav-vy and personnel to write software themselves when the need was critical.

The second wave of adoption carried HPC into industry, initially the automotive and aerospace sectors, starting in the late 1970’s. These users required HPC vendors to provide not only an operating system and other system software, but to port the key third-party ISV applications needed to run the users’ industry-specific problems — and to run the applications with the reliability expected in pro-duction computing environments.

The third important wave of HPC market growth began with the birth of the cluster in the late 1990s and took off in earnest in 2002. Since that year, clusters have almost

single-handedly driven the rapid growth in the HPC mar-ket and are now the dominant species of technical servers, representing about two-thirds of HPC systems sold each year. The compelling price/(peak) performance of clus-ters further democratized the HPC market, allowing even SMBs and SMSOs (small and medium-size science orga-nizations) to benefit from this game-changing technology.

And now it’s time to roll out the red carpet for another new wave of users: organizations that are migrating to HPC to perform high performance data analysis (HPDA), sometimes in near-real time, using graph analytics, semantic analysis and other advanced methods.

Examples of large organizations in this category include PayPal, which IDC estimates has saved over $700 million by adopting HPC for real-time detection of online con-sumer fraud; German pharmaceutical firm Schroedinger, which now screens massive numbers of drug candidates in the public cloud; and the U.S. Government’s Centers for Medicare and Medicaid Services, which has been evaluating how HPC can help to capture more of the $100 million-plus lost annually to fraud in the programs CMS oversees. Smaller first-time adopters come in many shapes and sizes.

Examples here include horizontal players such as Emcien (Atlanta, GA), whose cloud-based fraud detection algo-rithm is used everywhere from classified government to retailers; and vertical players like Apixio (San Mateo, CA), which applies graphing to heterogeneous medical data to assess risks, help manage patient populations, and bal-ance quality and costs for large providers.

IDC forecasts that thanks to new arrivals like these, along with organic growth, the worldwide HPC market will continue its healthy growth curve to reach $30 billion in 2017. It’s an exciting time to be part of this community.

Move Over, HPCers: Another Wave of Immigrants is Hitting Your ShoresBy Steve Conway, IDC Research VP, High Performance Computing/Data Analysis

• • • •

• •

•

Steve Conway

ANNIVERSARY25TH

How important is SC? I am always amazed how many people come to SC from around the world. Frequently the best new ideas (as seen in awards from Gordon Bell to Best Paper) are from around the world too. HPC is truly and international field and SC is the best place to see it all.

– Jeff Hollingsworth, University of Maryland

How important is SC to the global HPC community?

The global HPC market’s enormous expansion from about $2 billion in 1990 to $21.9 billion in 2012 has been heavily driven by successive waves of new users, with each new group of arrivals carrying its own expectations and opportunities.

“

“


DATE LOCATION SPEAKER TOPIC

1988 Orlando, FL Seymour Cray [Cray Research, Inc.] What's this about Gallium Arsenide?

1989 Reno, NV John Rollwagon [Cray Research, Inc.] Supercomputing – A Look Into the Future

1990 New York, NY Danny Hillis [Thinking Machines Corp.] The Fastest Computers

1991 Albuquerque, NM Allan Bromley [Chair, OSTP] The President's Initiative in HPCC

1992 Minneapolis, MN Larry Smarr [NCSA] Grand Challenges! Voyages of Discovery in the 1990s

1993 Portland, OR Neal Lane [NSF] HPCC and the NII

1994 Washington, D.C. Edward McCracken [SGI] Making the NII Real

1995 San Diego, CA William Wulf [Univ. of VA] And Now For Some "Really" Super Computing

1996 Pittsburgh, PA Frances Allen [IBM] Scaling Up

1997 San Jose, CA Paul Saffo [Institute of the Future] Is Digital Dead?

1998 Orlando, FL Bran Ferran [Walt Disney Imagineering] There's No Bits Like Show Bits

1999 Portland, OR Donna Shirley [NASA] Managing Creativity in Technical Projects

2000 Dallas, TX Steve Wallach [CenterPoint Ventures] Petaflops in the year 2009

2001 Denver, CO J. Craig Venter [Celera Genomics] Accelerating Discovery through Supercomputing

2002 Baltimore, MD Rita Colwell [NSF] Computing: Getting us on the Path to Wisdom

2003 Phoenix, AZ Donna Cox [Univ. of Illinois, NCSA] Beyond Computing: The Search for Creativity

2004 Pittsburgh, PA Tom West [National Lambda Rail] NLR: Providing the Nationwide Network Infrastructure for Network and "Big Science" Research

2005 Seattle, WA Bill Gates [Microsoft] The Changing Role of IT in the Sciences

2006 Tampa, FL Ray Kurzweil [Inventor, Author, Futurist] The Coming Merger of Biological and Non-Biological Intelligence

2007 Reno, NV Neil Gershenfeld [MIT] Programming Bits and Atoms

2008 Austin, TX Michael Dell [Dell, Inc.] Higher Performance: Supercomputing in the Connected Era

2009 Portland, OR Al Gore [Former U.S. Vice President] Building Solutions: Energy, Climate and Computing for a Changing World

2010 New Orleans, LA Clayton M. Christensen [Harvard Business School]

How to Create New Growth in a Risk-Minimizing Environment

2011 Seattle, WA Jen-Hsun Huang [NVIDIA] Exascale: An Innovator's Dilemma

2012 Salt Lake City, UT Michio Kaku [City Univ. of New York] Physics of the Future

2013 Denver, CO Genevieve Bell [Intel] The Secret Life of Data

SC 25th Anniversary – The Complete History of KeynotesCourtesy of The Exascale Report

• • • •

• •

•


You’ll find that most of the good bars and entertainment are in the LoDo area of Downtown. It’s a bit of a hike from the convention center, but you can hop on the free 16th Street Mall Shuttle bus and get off on Blake Street to catch the action.

Local’s Guide to Bars and Entertainment in Denver• • • •

• •

•

Local Brewpubs

Wynkoop Brewing Company


(303) 297-2700

Breckenridge Brewery 2220 Blake Street Denver, CO 80205

303-297-3644

Drinks with a View

Peaks Loungeat the Hyatt Regency Convention Center


303-436-1234

Jazz Music Dive Bar

El Chapultepec1962 Market Street Denver, CO 80202

(303) 295-9126

insideHPC Pub Crawl

The Celtic Tavern1801 Blake Street Denver, CO 80202

(303) 308-1795

Churchill’s Public House1560 Boulder Street Denver, CO 80211

(303) 477-1600

Vesta Dipping GrillMid-crawl snack

1822 Blake Street Denver, CO 80202

(303) 296-1970

Fado Irish Pub & Restaurant


(303) 297-0066

Howl at the Moon Denver


(303) 291-0880

Tilted Kilt Pub & EateryFor the Coyote Ugly fans

1201 16th Street Mall #120 Denver, CO 80202

(303) 893-5458

Fine Cigars

Palma Cigars Co2207 Larimer Street Denver, CO 80205

(303) 297-3244

Delaney’s Cigar Bar and Lounge

1805 Blake Street LoDo, CO 80202

(303) 308-1795

Spas

Veda Salon & Spa1717 Champa Street Denver, CO 80202

(303) 308-0524

Izba SpaRussian Banya Massage Therapy

1441 York Street Denver, CO 80206

(303) 321-1239


An Interview with Brian Sparks and Scot Schultz from MellanoxThe HPC marketplace can be very challenging for companies trying to make their mark. Time and again, Mellanox raises the bar in this area with global events and a keen eye on building community. We caught up with Scot Schultz and Brian Sparks from Mellanox for an insider’s view of what it takes to succeed.

insideHPC: How has marketing in HPC changed over the years you’ve been serving the community?

Scot Schultz: Marketing for HPC today, at least from the perspective of Mellanox, has evolved into more of an educational ex-perience. It’s not just about touting products and solutions on product specification sheets. We participate heavily in the community to drive an interactive level of educa-tion on why our products and solutions matter. We also participate with industry organizations such as the HPC Advisory Council to understand HPC use cases in-detail. Mellanox has even recently announced the Mellanox Academy, a new innovative learning experience — com-plete with access to Mellanox experts and flash-based in-teractive self-paced units.

Brian Sparks: To add to what Scot said, our participa-tion with the HPC Advisory Council is probably one of the most rewarding parts about my job. It really is a great feeling to help provide opportunities for students from around the world to participate in HPC competitions that showcase their skills and reward their endeavors in front of the HPC community and its leaders. In fact, some of

these students have quickly been noticed for their skills and have received job offers on the show floor. It doesn’t take a genius to see that new, young blood is critical to our industry. Much is to be done for us to get to exascale and beyond, and our biggest problem isn’t so much the hard-ware and software…it’s people. Say all you want about the rest, but attracting top talent to this industry and make it an attractive career for them is probably this industry’s toughest challenge.

insideHPC: Do you think marketing in this space is easier or more difficult than in mainstream, commercial computing markets?

Scot Schultz: Because HPC crosses so many technical disciplines, it’s more important that the end-users realize why a technology is an ideal fit for their particular envi-ronment, so it’s much more exciting -- as well as challeng-ing.HPC users have a very tailored set of requirements — from specific versions of their software stack, compilers

and math-libraries to applied acceleration technologies, and even storage. Typically, HPC and technical comput-ing users can’t afford to rip and replace, so it’s about mar-keting the technology benefits and how the technology is relevant to their needs today and tomorrow. Take for instance, a cluster that is built in phase I, that will be up-graded to included GPU’s for acceleration in phase II, the following year — so understanding the benefits of GPU-Direct RDMA, which is a no-cost software update, will further increase performance and decrease the latency of GPU-GPU communications in their environment; this is an example of marketing the value of an HPC solution, not just a product.

Brian Sparks: To me, HPC is much easier in the fact that you know who your target audience is on a personal level. In fact, it’s very hard not know everyone involved in any shape or form of HPC…whether it be an OEM, ISV, or End-User. That’s what I love about the HPC market, it’s

Brian Sparks and Scot Schultz on the Secrets of Technical MarketingInside Technical Marketing in the HPC Space

• • • •

• •

•

Brian Sparks

Scot Schultz

ANNIVERSARY25TH

How important is SC? It is regarded the HPC conference. Most of my colleagues ensure to be preset at every SC because it is our most important networking event. This creates a unique atmosphere for learning about ongoing research in the technical program or in private conversa-tions over a beer.

– Torsten Hoefler, ETH Zürich


Typically, HPC and technical com-puting users can’t afford to rip and replace, so it’s about marketing the technology benefits and how the technology is relevant to their needs today and tomorrow.

““


very easy to get to know everyone and understand what they want, as opposed to the millions of data center / com-mercial end-users.

insideHPC: What mistakes do you see companies make when launching marketing programs for HPC products?

Scot Schultz: Some companies just assume or expect the end-users will know how their product will be used, or more importantly that it could be applied to their environ-ment. But it rarely is about the product, it’s really about how your product fits into a solution! When a particular marketing program is launched, it needs to be inclusive of all the methods you might expect, such as a press re-lease, web seminars, print and web-banner presence, etc…, but also be inclusive with customer studies on how it’s technically being used, non-biased white-papers and of course training. Mellanox even offers its solutions for evaluation, at no cost to the customer, giving them the ability to evaluate our technology to ensure it meets their needs.

Brian Sparks: I wouldn’t go as far as saying people are making mistakes…it really comes down to budget and resources. Most of the companies in the HPC market are very small with limited staff and budget, especially on the marketing side. HPC is very R&D centric, and when it comes to marketing, I think most of these companies do what they can with what they have.

insideHPC: Do you think technology companies tend to underestimate the importance of marketing their wares?

Scot Schultz: Technology companies are usually driven by engineering, especially in their early stages of formation; so marketing isn’t always the primary concern. However, a well-balanced organization will certainly value the pow-er of clear messaging on all aspects their products bring to the market. Just because a product addresses a void or some need in the market, doesn’t make it a value — you need to be very clear and demonstrate exactly why one should care about what you have to offer. A well planned marketing campaign will make a tremendous impact on interest and demand generation — technology companies generally realize this, but the ones that don’t get it aren’t around for too long — no matter how clever their technol-ogy might have been.

Brian Sparks: Not to repeat what I said above, but the majority of HPC-centric companies are very small and are in a start-up mentality. It basically means the engineers of the company must also be marketing-savvy and pull double duty, otherwise traction in the market, as Scot said, will not be performed as quickly as they would wish. You may have a great product, but will anyone hear about it? Not unless you have strong end-user advocates who are willing to help market it for you through word-of-mouth or social media.

insideHPC: How much does marketing really help if an HPC vendor’s target prospects are the government labs?

Scot Schultz: One thing is for certain, the government labs employ some extremely savvy technical experts. Another thing you can be sure of, those experts are still human be-ings, and desire their technical skills to be kept razor sharp. Learning is a constant endeavor — Technology changes, and how one technology might be applied to adjacent innovations is constantly evolving as well. Cloud-based computing for HPC workloads is a good example of this, which is being explored by the labs today — new innova-tions in this area are happening and Mellanox’s SR-IOV solutions will play a key role in achieving real bare-metal performance in these environments… so for Mellanox the marketing/education for those who work in the govern-ment labs is really not that different.

ANNIVERSARY25TH

My favorite year? SC12 since I was general chair. It was like hosting the biggest most expensive party I have ever thrown. By the end I was twice as exhausted as a normal year, but didn’t want it to end.



ANNIVERSARY25TH

My fondest memory of SC? It was my first Beobash. No, I don’t remember precisely when. I do remember arguing strongly within SGI about these Linux clusters that had po-tentially very disruptive potential to our core market, spe-cifically that we should get out front of them and lead with them. Beobash was nice in that it brought a bunch of like minded people together to talk informally over drinks.

– Joe Landman, Scalable Informatics


You may have a great product, but will anyone hear about it? Not unless you have strong end-user advocates who are willing to help market it for you through word-of-mouth or social media.

“

“


Brian Sparks: Marketing to government certainly adds a little complexity. What you did on the commercial side may not be okay on the Government side. For example, giveaways at shows are pretty much a no-no, so you re-ally need to change the way you get government labs to remember you. To me, it’s talking with them at their level. Remove all of the marketing BS and stick with the techni-cal facts and get across your points on why you believe your solution is a match to their initiative/project.

insideHPC: How important is it to maintain a balanced inter-national approach to HPC marketing programs?

Scot Schultz: HPC and technical computing has no boundaries, even more-so today it is not only used by nearly every country, information and data is exchanged on a global scale. Humanity is naturally driven to resolve understanding of key issues; like global warming, earth-quake and severe weather prediction, and finding cures for illness to name a few. Awareness for new technology and HPC advancements is just as important to even the smallest nations of the world.

Brian Sparks: Scot is absolutely right. Going back to our work with the HPC Advisory Council, the Student Cluster Competitions are very much a global competition. Apart from the usual suspects, we have had teams from South

Africa (who was the winner at the HPCAC-ISC’13 Student Cluster Competition), Costa Rica, and for 2014 we expect a team from Brazil to enter, marking the first time a South American team has entered the competition. I think you will always see a weighted approach in terms of market-ing programs targeting North America, Europe and Asia, but more and more companies should be looking in the southern hemisphere if they aren’t already.

Visit us to learn more at Booth #3518

See how teams of international undergraduate students race to demonstrate their greatest performance in designing their own cluster. Join us in Leipzig, Germany to witness the best and the brightest on the exhibit floor during ISC’14.

ANNIVERSARY25TH

How important is SC? One downside of HPC is that we tend to be scattered everywhere, with just a few people at each organization. SC is *the* place to meet up with other people involved in similar HPC work, from around the globe. I can’t tell you how many of my own research collab-orations have started through meeting someone at SC who was interested in similar problems. That’s why I’ve been a faithful participant ever since my first visit at SC89. I get so much out of the networking opportunities.




As thousands of global HPC stakeholders prepare to as-semble in Denver for the 25th anniversary of the SC con-ference, SC13, the staff of The Exascale Report is thrilled to present an upbeat discussion on the topic of Exascale.

Over the past three and a half years, we’ve tried to present a balanced perspective of community opinion by covering both the positive and negative discussions of exascale.

We’ve praised the worthwhile efforts and creative spirit of many community leaders, while also stepping up to ‘tell it like it is’ when politics or lack of leadership were threaten-ing to bring an entire ecosystem to a dysfunctional halt.

In hindsight, perhaps we all expected too much.

In The Exascale Report article from July 2010 titled DARPA and UHPC: Jump Starting a Revolution, (written by John Kirkley), we noted that DARPA was looking for a 1,000X increase in computing capabilities by 2015. In that same issue, we had numerous references to the widening chasm between two camps on what it would take to get us there. With much praise from our readers, the Exascale Report placed a spotlight on the discussion of Evolution vs. Revolution.

That was then. This is now.

Here we are, almost three and a half years later, and it doesn’t look like we’ll hit that 1,000x performance im-provement milestone by the year 2015. Most industry ex-perts feel that 2020 or 2022 is a more realistic timeframe. And truth be told, there are still arguments, albeit mostly semantic, around the discussion of a revolutionary vs. evolutionary approach being the best path to take us to exascale-level computation.

The HPC Battlefield Flashback: 25 years of SC conferences

I remember well my first exposure to the HPC community. I was working with the VLIW company, Multiflow Com-puter, when I attended Supercomputing ’88 in Orlando. There were only 1,400 attendees that year, (as a contrast to the size of SC today), but I was hooked on HPC from that first event. One thing that I observed back then — and every year since then, is that this has never been and likely never will be a community lacking in ideas — or opinions.

Over the past quarter century, Gallium Arsenide and CMOS prophets took strong positions, as did the opposing camps of SIMD and MIMD, Open Source and Proprietary, RISC vs. CISC, NUMA vs UMA, Vector Supercomputers vs. Killer Workstations, and Supercomputers vs. Clusters.

These battles often rolled into the technology media and became key discussion points at the SC conferences, go-ing back to the very first keynote in 1988 when Seymour Cray’s presentation was titled, “What’s All This About Gallium Arsenide?”

Progress in the HPC space has been steady and impres-sive, but unfortunately, has created an environment of high expectations. I think at times, it’s difficult for some of us to see and appreciate important stepping stones of progress when they aren’t wrapped in holiday paper and presented as major milestones.

An Update to the Exascale Progress MeterHPC Community Says We Are Moving in the Right Direction

Exascale Report to Recognize One Community Leader as a Vanguard of HPCBy Mike Bernhardt

• • • •

• •

•

It’s difficult for some of us to see and appreciate important stepping stones of progress when they aren’t wrapped in holiday paper and presented as major milestones.

“ “

ANNIVERSARY25TH

My fondest SC memory is my first SC paper presentation in 1992. I was a graduate student and excited to give a talk to a big audience. They had told us we would have two over-head projectors so my talk was prepared with slides that need to be moved from one projector to another at specific points. I get to the room for the talk and find the two pro-jectors were in opposite corners about 75 feet apart. I am panicked about how I can give my talk while running back and forth. Eventually I get a “volunteer” to move the slides between projectors for me at the correct times. Years later I realize the experience was a great lesson about HPC, the hardest part is getting the data to move to the right spot when needed!




The 2013 Reality Check: It’s not unanimous, but it is a majority opinion. We are making progress.

A small number of leaders in the HPC community rate our 2012-2013 progress toward exascale as poor. They cite lack of leadership, inadequate funding, and no solid direc-tion. But based on our recent survey, this view is in the minority.

The results of our recent Exascale Progress poll tell a dif-ferent story. Taking an average score from almost 400 sur-vey respondents, the vast majority of HPC stakeholders believe the Exascale Progress Meter has moved two points — and in the right direction.

This is indeed good news. The emerging exascale commu-nity has spoken. We are making progress.

It is worth noting that a number of respondents cited their optimism around the appointment of our new Secretary of Energy, Ernest Moniz. His support of exascale beyond some initial statements is still to be seen, but many com-munity leaders seem to have a renewed confidence that we have a supportive voice on the President’s cabinet.

Using the SC Conference as our frame of reference, we’ve had significant challenges over the past two years as Washington politics directly impacted the productivity of our nation’s scientists and researchers — cutting budgets,

eliminating travel, and creating environments that made collaboration a painful process. We’ve had concerns for the past several conferences regarding attendance at SC, but for the most part, it was unfounded. Department lead-ers were able to get a respectable number of key research-ers and scientists to the conferences — through dogged determination and creative accounting.

Through these very difficult times, the HPC community held strong and pushed forward, driven by passion, deter-mination, and the deep-rooted understanding that scien-tific discovery and economic progress can’t simply be put on a back shelf.

If the HPC community has been able to thrive, even con-servatively, during the most dysfunctional times, imagine what we can do with strong leadership. Realistically, we may not be able to convince Congress or the White House to allocate the level of funding that is necessary to drive a national technology agenda at a competitive pace, but this community, the emerging exascale community, will push forward. We are making progress.

The real core of the HPC community is the people — and this is a many, many-core community.

With our first exascale progress meter back in 2011, the community rated our progress as 5.5 on a scale from 1 to 20. Most people voiced frustration aimed at funding agency leadership, along with confusion around an exas-cale roadmap and which path to take.

Today, that is different. While a few people still feel that way, the majority of the community is encouraged — largely because of the new DOE leadership, but also be-cause of the tireless efforts of a few organizations and key individuals.

Join us for our Special Evening Event during SC13

Interconnect Your Future

Keynote by Eyal WaldmanPresident and CEO, Mellanox Technologies

With guest speakers from IBM and NASA

www.mellanox.com/sc13

REGISTER TODAY

November 20TH 7PM - 10PM

Connect with us:

Visit us at SC13Booth #2722

ANNIVERSARY25TH

How important is SC? Quite important as a meeting place where high bandwidth, high density, close proximity net-working (people) can be accomplished, which is often harder to do when users/partners are spread out all over the world. Given the recent massive increase in interest in massive scale analytics (e.g. “big data”), which leverages HPC concepts, the relevance of HPC in the global comput-ing community is more so than in the past. Having these meetings, and being able to extend them to cover topics of direct relevance to many others is important to keep ideas and information flowing. Which helps everyone.




Our poll reflects the opinions of almost 400 users (non-vendors) throughout the HPC community. On our scale of 1-20, scores ranged from a low of 3 to a high of 13. The average came out to 9.5. That score equates with ‘good’ progress.

The general perception of our forward progress on the path to exascale has improved 4 points since our first poll in 2011.

When we conducted our recent poll to determine com-munity perception of exascale progress, we also asked our readers if any organizations or individuals stood out and deserved to be recognized for their leadership and contri-bution to keeping the HPC community strong and moving forward. We solicited nominations to recognize one orga-nization or one individual as a Vanguard of HPC.

The response was extremely positive. We received 25 nom-inations from almost 400 people. Of those 25 nominations, six individuals are in a very tight race to see who will be the recipient of the inaugural HPC Vanguard Award.

These six have already risen to the status of legends in HPC, but only one, as determined by a vote of HPC com-munity peers and colleagues will be selected for the recog-nition of HPC Vanguard.

The Exascale Report will be announcing the winner of the first HPC Vanguard Award at SC13 on Monday, Novem-ber 18th.

We expect to see a prototype exascale operating system sometime in 2016. We have new leadership at the helm of the U.S. Department of Energy. There is renewed enthu-siasm and hope among many HPC stakeholders, and we anticipate a very exciting SC13 conference.

So, as thousands of us gather at the Colorado Conven-tion Center and the many fine watering holes of Den-ver to celebrate the 25th Anniversary of the SC conference, I feel there is a lot for which we should be thankful. We are making progress.

Community Recognition

The six finalists the HPC Vanguard Award — individu-als who have been nominated by a community vote for having played a key role in moving the exascale progress meter forward, are:

• Pete Beckman, Argonne• Bill Dally, NVIDIA• Jack Dongarra, University of Tennessee• Alan Gara, Intel Corporation• Bill Gropp, University of Illinois

(Champaign-Urbana)• Thomas Sterling, Indiana University

The winner of the inaugural HPC Vanguard Award

will be announced at SC13 at a press briefing on

Monday, November 18, 2013 at 3:00PM

ANNIVERSARY25TH

My favorite year? For me, possibly the most affirming con-ferences were the ones where accelerators broke out hard into the mindset of the community. This is in part due to what we were trying to do in 2002-2006 ... its nice to see that our predictions (and heck, the whole business model and ecosystem) evolve right about the way we predicted.



2011 Progress: 5.5 2012 Progress: 7.5 2013 Progress: 9.5

None

L

ittle

Little Good Good Im

pressive

Very

Very

None

L

ittle

Little Good Good Im

pressiveVe

ry

None

L

ittle

Little Good Good Im

pressive

Very

The Exascale Progress Meter (2011 – 2013)

VeryVery


Denver is the capitol of Colorado, and with a population of 619,968, it is the 23rd most populous city in the US as of 2011. The western state of Colorado is about 1,800 miles from New York on the east coast and just over 1,000 miles from Los Angeles on the west coast.

Nicknamed the “Mile High City” due to the fact that its official elevation is exactly one mile (5,280 feet) above sea level, it has one of the highest elevations of all the major cities in the US. Denver is situated in the South Platte Riv-er Valley, just east of the Rocky Mountains. Home to sev-en professional sports teams, including the Denver Bron-cos, the Colorado Rockies, the Colorado Rapids and the Denver Nuggets, it’s an ideal place for the sports fanat-ic no matter what time of the year it is. For this reason, www.denver.org rightly refers to the city as a “pro sports paradise”. The weather in Denver is very inviting. Winters are remarkably mild with an average high of 45 degrees. It does get cold enough to snow at times, but it rarely stays on the ground for very long. If you’re a golfer, you’ll be happy to know that Denver has plenty of courses, and they are typically open all year round.

In 1858 during the Pikes Peak Gold Rush, Denver was founded by General William Larimer Jr. on the St. Charles Town Claim as a mining town. Under threat of being hanged, a representative of St. Charles Town Co. agreed to give up his claim in exchange for a barrel of whiskey, which may not seem like much, but considering the alter-native, it was an offer he couldn’t refuse. This land was an unlikely place for a city to be founded. Throughout its early history, many attempts to build cities in Colorado re-sulted in failure, which is why there are about 500 “ghost towns” in Colorado today. A ghost town is a town, city or village that has been completely abandoned usually due to either natural disaster or economic failure.

Denver could have very easily gone the way of these ghost towns due to a number of factors. It was not built on an existing road or railroad, neither was it built on a lake or river or near a body of water. By all accounts, this was an unfavorable location for a city. In the words of travel writ-er Rose Kingsley, “It was as if the angels were carrying a city to a proper place and accidentally dropped it here.” In fact, in its early history, Denver was destroyed and re-built twice, once as a result of fire, and once as a result of floods. Yet despite the odds, Denver ultimately survived and thrived. To a large extent, Denver owes its survival to ambitious people like William Byers who helped bring the railroad to the city and who in 1859 founded the Rocky Mountain News which was used to promote this new city and give it some credibility in the eyes of the rest of the nation. Or John Evans who not only built railroads and

churches, but also built the University of Denver which was founded in 1864. Those are just a couple examples of the entrepreneurial spirit of the early citizens of Denver that ensured that this city did not become just another ghost town.

Over the years, Denver has managed to evolve while still retaining its historic, old west roots. To this day, snapshots of its early history have been preserved and are proudly displayed throughout the city. For example, it is home of the renowned Buffalo Bill Museum and Grave, which was named best museum of the year by True West Magazine in 2010. William Frederick Cody, AKA Buffalo Bill, was an American soldier, bison hunter and showman who lived from 1856 to 1917. Before Buffalo Bill died, he requested that he be buried on Lookout Mountain which overlooks the Great Plains and the Rockies and is located about half an hour from downtown Denver. In 1921, the Buf-falo Bill Museum was constructed at the site of his grave by Johnny Baker, and since then, it has attracted millions of visitors and is one of the top tourist sites in Colorado. The museum is a great spot for history buffs and offers an informative and entertaining experience, featuring an ob-servation deck, a gift shop, a café, hiking trails and more.Admission is very affordable at $5.00 per person, $4.00 for seniors, $1.00 for children between the ages of 6–15 and free for kids 5 and under. During the winter months, the museum is open to the public Tuesday through Sunday from 9AM – 4PM.

Another one of the top tourist sites in Denver is the Den-ver Performing Arts Complex, also known as the “Denver Center”, located on 13th street in downtown Denver. This is one of the largest performing arts centers in the country, second only to the Lincoln Center in New York City. The four block, twelve acre site holds ten performance spaces with a combined seating capacity of over 10,000 people. The performance spaces include the Ellie Caulkins Opera House, the Buell Theatre and the Boettcher Concert Hall.

SC13 Comes to Denver, Coloradoby Robert Murphy

• • • •

• •

•

ANNIVERSARY25TH

My favorite year? I’d have to say SC99. That year I was privileged to serve as conference chair, which gave me a really unique opportunity to work with all areas of the con-ference and interact with all our different audiences. It was an intense experience, but extremely rewarding both pro-fessionally and personally. (I suspect any other former chair that you ask will say something similar!)



http://www.denver.org


Here you can be treated to a wide variety of performances, including Broadway touring productions, opera, dance and ballet, a major symphony orchestra, a Tony Award winning theatre company and more. For information about tickets and upcoming events, check out www.den-vercenter.org.

Denver also has a great zoo, which in 2005 was named the most popular paid attraction in the Denver metropolitan area. The 80-acre zoo is accredited by the association of zoos and aquariums (AZA) and is a member of the world association of zoos and aquariums (WAZA). Alexander J. Graham began the zoo in 1896 with an orphaned Ameri-can black bear cub. It has come a long way since then and today is home to 3,500 animals from over 650 different species. The zoo is known for its authentic, naturalistic habitats which not only serve to enhance the experience for zoo patrons, but makes life happier for the animals as well. The Denver zoo is also committed to environmental sustainability by means of water and energy conservation, alternative transportation and local food and is currently working towards becoming a zero waste facility by 2025. From November through February, admission is $12 for visitors ages 12–64, $10 for ages 65+, $8 for children ages 3–11 and free for kids 2 and under. Group discounts are also offered. During the winter, admissions are open from 10am – 4pm with grounds staying open until 5pm.

On the corner of West Colfax Ave and Delaware St you can find the Denver Mint, a branch of the United States Mint. You might want to consider taking advantage of the opportunity to take a free tour, as the Denver Mint is one of only two US mint locations (along with Philadelphia) that offers tours to the public. In 1860, during the Pikes Peak gold rush, Clark, Gruber and Company began to pro-duce coins using gold dust. In 1863, the facilities of Clark, Gruber and Company were acquired by the US Treasury for $25,000 and the United States Mint of Denver officially opened as a United States Assay Office. In 1906, the Den-ver Mint was moved to its current location on West Colfax and Delaware and for the first time began to mint silver coins in addition to gold ones. The Denver Mint has been an important part of Denver’s history and of the history

of the United States. Its founding goes back almost to the founding of Denver itself and to this day it is the largest producer of coins in the entire world.

The 16th Street Mall is yet another extremely popular tour-ist spot in Denver. The pedestrian/transit mall opened in 1982 and originally went from Market Street to Broad-way. In 2001, the mall was extended and it now stretches for over a mile along 16th street in downtown Denver from Wewatta Street to the intersection of 16thAve and Broadway.A free shuttle bus service known as “MallRide” runs the entire length of the mall, stopping at every in-tersection along the way and also connects to the Denver Light Rail. MallRide operates 7 days a week. The mall is home to the Performing Arts Center, the Colorado Con-vention Center, a comedy club, hundreds of stores and shops, not to mention over 200 restaurants including ev-erything from fast food to fine dining to coffee shops. VIS-IT DENVER’s Visitor Information Center is also located in the mall on the corner of 16th and California. In addition to all of this, on any given day at the 16th Street Mall, you will encounter a colorful conglomeration of street per-formers and vendors.

Public transportation in Denver is affordable and conve-nient. You can get almost anywhere in the city via bus and/or light rail. Beginning on April 26th of this year, Regional Transportation District (RTD) will be adding the West Rail Line, an extra 12.1 miles of light rail running be-tween Denver, Lakewood and Golden on the W line, mak-ing public transportation in the Mile High City that much more efficient. For information about fares, bus lines and schedules log on to www.rtd-denver.com. Another avail-able option for transportation is Yellow Cab of Denver www.denveryellowcab.com which is not only convenient, but can also be an environmentally friendly way of com-muting now that they have added hybrid and propane cars to their fleet.

As you can see, Denver has quite a bit to offer in the way of culture and entertainment. This combined with the pleasant weather and reasonably priced hotels and variety of dining options should make for a very enjoyable visit to Denver.

BE SAFE.

One thing you should note when visiting Denver — the crime rate in Denver is significantly higher than the national average.

Be aware of this and use common sense. As a general rule, it is probably not the best idea to walk around the city at night if you are not in a group. Keep that advice in mind and have a safe and wonderful time exploring the Mile High City.

The 16th Street Mall is home to the Performing Arts Center, the Colorado Convention Center, a comedy club, hundreds of stores and shops, not to mention over 200 restaurants includ-ing everything from fast food to fine dining to coffee shops.

“

“

http://www.denvercenter.org

http://www.denvercenter.org

http://www.rtd-denver.com/

http://www.denveryellowcab.com/


Minute as its impact may be in our physical uni-verse, the fact of quantum entanglement is this: If one logically inexplicable thing is known to exist, then this permits the existence of all logi-cally inexplicable things.

– Brian McGreevy, Hemlock Grove

The Accident 6:02 am PDX Airport

Jake straightened his badge as he walked up to the security checkpoint. The TSA supervisor would be onsite already as always, but it wasn’t about that. Jake liked to look sharp and the dream he had last night had rattled him. He was sure that his coworkers would see it on his face.

It didn’t help that the dream was still so vivid in his mind. It had taken place right here at work, only something in the dream was different. A feeling more than anything, Jake still carried with him the overpowering sensation that he was being watched.

As he took his seat at the body scan-ner controls, Jake wondered if he had made a mistake downloading the Shad-ow App. It was something new, and very exciting — technology that enabled people to remem-ber their dreams. As instructed, he recorded them into the worldwide database. It was so simple after all, just tap-ping a few characters into his smart phone in the middle of the night. But as he continued using Shadow, the more ef-fective it became. Dreams no longer faded. Instead, he car-ried them around with him all day — even the bad ones.

“You look tired, Jake,” said Father Joseph as he entered the checkpoint with an extra cup of coffee in his hands. “Care for a Hair Bender Americano?”

A gaunt man in his sixties, the priest had struck up quite the friendship with Jake over the past few weeks. He liked his Stumptown coffee from downtown, and by bringing Jake an extra, he was able to smuggle in his own cup each day on his way to the meditation chapel at C Gate. They would often talk over coffee during breaks, and Jake liked to call him Joseph, the Explainer of Dreams.

Jake was new to the agency. After a couple of tours as an MP in the Balkans he had wandered from job to job, moon-

lighting as a security guard at the docks. It was mostly solitude there, a time of healing for him after accidently shooting an unarmed man in Serbia.

The man had stolen bread from the PX and ran from Jake while he was on patrol. In the dark, the pointy baguette in the man’s hand looked like a knife.

Now Jake took the coffee from his friend and set it on the console. He could use the caffeine after his rough night.

“Same dream again?” Father Joe asked.Jake nodded; he raised his hands to describe the box he

had held in the dream, the one with the sounds of breath-ing inside. As he did, his arm knocked the cup of Stump-town onto the console.

“Dammit!” Jake exclaimed, wiping up the coffee with his handkerchief. “I’ll be screwed if I messed this thing up.”

“I’m sorry, Jake. We can talk later. If you’re ready to test it, I’d like to get scanned and get on over to the chapel now.”

As he did each and every day, Father Joe raised his hands above his head as the scanner did its business.

Jake looked at the display absently. The monochrome readout was always the same, but not today. Father Joe’s scan had shifted red. No concealed objects, just red.

“This scanner is hosed,” Jake said out loud. “I better test it on myself.” He mo-tioned his colleague Jenkins across the checkpoint to take his station and then he entered the scanner. It whirred around him.

“Looks normal to me,” said Jenkins.Jake walked around to the display to take a look. “That’s so odd,” he said. “You ever see colors come out

of this thing?”Jenkins shook his head and returned to his post as Jake

watched the priest make his way to C gate. Redshift, he thought to himself. That’s what you see

when something’s moving away.

The Heist 6:20 am Elephant’s Deli

“Security is all about degrees of belief,” Pritis told his son as their car entered the parking lot. “Look at that sign: “Premises under 24 Hour Surveillance.” They think they are safe, but they are wrong. Your grandfather kept tell-ing me this after he lost everything in the Balkans. People always want themselves and their stuff to be safe, but they can never really truly have it.”

Sci-Fi Original: The Observer Effectby Rich Brueckner

• • • •

• •

•


“No,” Papa said to me. “They can only claim they have it.”Kotas was already taller than his father. At 16, he had

also developed into quite the accomplished thief as well. It was still dark and the boy’s black clothes and hoodie were the perfect cover for today’s smash and grab.

Pritis parked the Subaru in the back corner of the lot and pulled the jammer remote from his pocket. “OK, I’ll switch on the jammer and you hit those cars over there. I can’t leave it on for too long as this thing has incredible range. Your cousin in Russia got quite the bargain on this, I can tell you.”

Pritis activated the jammer and kept a close eye on his watch. Two minutes. That’s all young Kotas would need to break four windows and grab the baggage in the back of each car. The jammer would scramble the Wi-Fi cameras watching the lot. It shorted-out car alarms as well, with something his cousin called a “Rotating EMP pulse.”

He opened the hatch for Kotas and the boy ran up and loaded in the stolen gear. They then drove out of the park-ing lot, quiet as a cat in the night.

“Dad, you never told me what really happened to Grandpa in the Balkans. I think I’m old enough now.”

Pritis sighed. The boy was right. “I will tell you straight out. He was shot by an American for stealing bread. The war had destroyed everything. My mom and I, we were hungry and he went out.”

Pritis felt his voice trail off and they continued on in si-lence. As he drove, Pritis thought about the dream he had the night before. He was young again, a small boy waiting for his father to come home with some food. As he stared out the window and watched the long dirt road that lead to the house, gold coins starting raining from the sky.

The Escape 6:22 am Good Sam Mental Ward

Garth Meeks awoke suddenly from his dream. It was still dark outside and usually the meds kept him out until breakfast time at 8:00. It was always the same routine here in the ward, but not today. Garth was scared today, scared for his brother, Justin.

“I have to make a phone call,” he said to an empty room.Garth stood up and walked to the door. It was always

locked, but not today. Today he walked right through, past the reception desk and out the door into the pale be-ginnings of morning light.

He was hungry and his bare feet were cold on the pave-ment, but the deli next door was closed. Best to keep mov-ing, he thought. I have to make a call.

The Dark Gravity Experiment 7:30 am Cloud Labs International

Justin Meeks nodded to the guard and greeted Dr. Brandt at the door. “You’re right on time, Doctor. I’ve re-ally been looking forward to meeting you in person.”

Brandt shook Justin’s hand with great enthusiasm. “I could hardly sleep last night, Mr. Meeks. Are the super-computers fully configured?”

“Yes,” said Justin. “We’ve got the 512-Qubit quantum machine at Ames allocated for the entire day. It’s been a bear to deal with all the preprocessing data you supplied, but we’ve spun up an AWS instance with 50,000 cores to feed the beast. It’s more than we’ve ever done before.”

Justin led Dr. Brandt to the lab, which was decked out with dozens of displays. He showed Brandt a status dis-play of the company’s cloud software, a marvel of technol-ogy that automated the process of turning thousands of cloud servers into a virtual supercomputer.

Brandt was noticeably impressed. “This is a veritable mission control,” he said. “I was very pleased to learn that my experiment won the contest, especially considering the quality of the other proposals you received.”

“Yes, we were very impressed with your algorithms, Dr. Brandt. My CTO tells me your work here today will be a quantum leap for cosmology.”

“It’s more than that, Mr. Meeks. When we complete this simulation, we will be able to compare it to measured results that have been sealed away and encrypted in this memory key.”

The look on Justin’s face told the doctor he did not un-derstand.

“It’s rather complicated, but please let me explain. Are you familiar with the Quantum mechanics and the Ob-server Effect?”

“You mean, where the act of observing has an effect on the outcome?”

“Precisely. Only in this case, what we will be proving, or disproving, is the presence of an observer outside of our universe.”

Justin thought hard about what the Doctor was say-ing. “But I thought your experiment had to do with Dark Gravity.”

“Ah yes, yes it does. You see, gravity is a great mys-tery to us scientists. Out of the four primary forces, it is by far the weakest. Lucky for us, yes, because without weak gravity, the universe would have collapsed upon itself long before life itself could have evolved. It’s almost as if it was designed that way for a purpose.

“You’re saying that the universe conspired to create intelligent life?”

What power at a distance,” he thought to himself as the street lights went out around him. “This device is almost like a shadow that turns on its source.

“ “


“I’ll get to that. Now, as for gravity itself, obviously we can feel its effects, but graviton particles themselves have eluded us. The question is, why? This experiment hypoth-esizes that the full strength of gravity is an effect that we cannot observe directly because it is outside of the mem-brane of our universe.”

“Through the massive computational power and mem-ory of you’re quantum supercomputer, I will be able to run a complete simulation of the Big Bang, including the forces of Dark Gravity from an external membrane. And when I compare these results to the recently measured re-sults from the Planck satellite, I will have my answer."

“The answer to what, exactly?”“Well it’s very simple, Mr. Meeks. If this experiment

proves that there is an Observer Effect, then I intend to prove that there is an observer.”

Justin felt his knees going weak. The weight of what this man was saying came crashing down on him as Dr. Brandt continued.

“My dear Mr. Meeks. Today, this day, you and I and this mighty machine you have conjured up will prove, or disprove, the very existence of God.”

The Visitor 8:03 am Home of Pritis Admonics

Pritis grabbed the gym bag from the car and entered his house through the side door. The bag was heavy, as the fence was unwilling to take most of the booty from the car heist this morning. Dropping the bag on the kitchen table, he let out a sigh with the notion of having to sell the rest of this stuff to the invariable idiots that bought such things Craigslist.

Pritis looked around the kitchen for moment. Some-thing wasn’t quite right. He heard a sound from the den. Someone was in the house.

Almost instinctively, Pritis pulled the switchblade from his pocket. It was a nasty blade, sharp and serrated on one side.

“Who’s there?” he asked of the darkened room.Garth stepped out the shadows, his bare feet cold on

the hardwood floor. Still in his nightclothes, he regarded the knife in Pritis’ hand as something he thought he ought to recognize. Instead, his drugged mind instructed him to blink twice and return to the task at hand.

“I have to make a phone call.”“You’re looking for aphone?” Pritis replied, seeing the

dullness in the man’s eyes. “Idiot, why are you looking for a phone in my house!”

Garth stared straight ahead and felt his balance shifting to the left, like part of him was on some kind of boat.

“I have to make a call. I have to call my brother Jus-tin before he opens the vault. I saw it. His computers will make gold fall from the sky.”

This man was obviously sedated and not a threat. Pritis folded up his knife.

“Do not worry, my friend. I will get you a phone, but first, we should talk about your brother. Now tell me, where is this vault you speak of?”

Redshift 7:56 pm PDX Airport, Checkpoint Charlie

Jake arrived early for the extra shift he picked up from one of the other Agents, a nice lady named Marge who was fighting the flu. He would be taking over the scan-ner from Mason, kind of a squirrelly kid with thick glasses and bad skin.

“You’re early, Jake,” the kid said. “That’s awesome. It’s been a hell of day, what with happened to the priest and then this crazy machine and all.”

“Wait. What did you say about the Father Joe? Is he okay?”

“Oh, I’m sorry. You two were friends, I know. He passed away in the chapel this morning. It looked like a heart attack or something because he was slumped over, still kneeling at the altar when they found him.”

Jake felt the shock of his friend’s death wash over him for a minute before he could ask more questions.

“Is he gone, kid? Did they haul him out of here?”“Yeah, it was quite the scene. Cops from the Port Au-

thority. The coroner. The whole bit. And it was a hell of bad time for this scanner to go nuts. I mean, the line was backed up already half way to Starbucks and then all the scans for gate C1 were coming out red all of a sudden.”

“Red scans you say? Is it working now?”“Thank goodness, yes,” said the kid. “As soon as the

passengers started arriving for the C3 gate, the thing just started working like normal again.”

“You’re saying all the scans from C1 were red?”The kid nodded and returned to the scanner controls for

the last hour of his shift.Jake thought of the blurry red scan of his now deceased

friend. It was a Redshift, like something was moving away.Now he felt a rush of adrenalin as he made his way

quickly to Gate C1. Soon his panic lead him to a full run. If he was right, there might still be time to stop the plane.

As he rounded the corner, C1 was empty except for the ticket agent. Catching his breath, Jake walked to the win-dow and felt his hand reach out to touch the glass. It was just in time to see the Sun Gold Airways jet streak off into the sky.

It was time to turn on the jammer, which then blinked away as it acti-vated. Everything within a line of sight was going dark.

“ “


The Phone Call 8:05 PM Cloud Labs International

Justin got a text from his head engineer that the simu-lation run was nearly complete. He should have felt ex-citement, but his conversation with Dr. Brandt about the nature of the experiment left him with a creeping malaise.

As he was making his way down the hall toward the lab, his phone rang.

“Justin? Justin, are you there? This is your brother.”“Garth? Garth, what are you doing? You always call on

Sundays. What’s wrong?”“Justin, I’m with the police. They helped me make the

call. I have something, I mean, I have to warn you before it’s too late. The lady in white woke me up this morning. She told me your computers are going to open the vault. You need to go. Gold will fall from the sky.”

“Garth, where are you? Where is the police station??“Wait. Justin, they say I’m at the Burnside station on

Sixth. Please come now. I’m really scared.”Justin reached the end of the hall and found Dr. Brandt

waiting for him. He told the man he would be right back, but the Doctor barely seem to hear him and continued to stare at the progress display of the cosmic simulation.

The Shadow Turns 8:15 pm Outside Cloud Labs International

Pritis parked his Suburu in the shadows of an oak tree across the street from Cloud Labs. As he exited the vehicle, the noise from the nearby airport was nearly deafening. He hated such noises, as they reminded him of huddling in the cellar as a boy during the sorties of the Balkan war.

The roar of the jet began to subside now as he looked up to Western sky to see the Sun Gold airliner banking away in its steep ascent. Walking around the back of the build-ing, he surveyed the loading dock for WIFI cameras. Yup, there were at least three of them trained on the parking lot. It was time to turn on the jammer, which then blinked away as it activated. Everything within a line of sight was going dark.

“What power at a distance,” he thought to himself as the street lights went out around him. “This device is al-most like a shadow that turns on its source.”

Now all he had to do was jimmy the door and see if that escaped lunatic at the house was right about the vault.

The Observer Effect 8:25 pm Cloud Labs Mission Control

The blinking cursor on his laptop told Dr. Brandt that the simulation run was nearly completed. What he thought would take five minutes ended up taking ten, and he found his mind wandering back to the dream that fostered this experiment.

He was in a room alone. The room was empty except for a small wooden box on a table. He very much wanted to open the box, but he couldn’t move. When he closed his eyes, he found he was able to move his hands. He reached out in the darkness and touched the latch on the box. Eyes still closed, he opened it now. He could feel what was in-side, but he couldn’t identify it. And as he was about to open his eyes and see the contents, he heard the sound of breathing. Someone was behind him. Suddenly terrified, he turned and forced his eyes open to the golden morn-ing light that streamed through his bedroom window. He looked up to see that his hands were raised above the bed as if he was still holding something, something that had changed when it was seen.

Now his awareness returned to the lab as the lights in the lab went out. It was dark except for the glow of his laptop screen.

Now he heard the sound of something scraping on metal. Someone was entering through the steel fire door that lead to the loading dock.

A hooded man entered the lab carrying a small black remote control in his hand. Brandt found himself at loss for words as the man approached him.

“You opened the door,” Brandt offered in observation.“Yes,” said Pritis, looking around the room for a safe.

“Have you?”Brandt heard the ping from his laptop that signaled

his post-processing job had completed. No time for the stranger, he looked down to see the simple binary results of the great experiment.

He felt his breath escape as he froze there in that mo-ment of knowing. In all the faithful, in all the history of the world, there could be no more degrees of belief.

Puzzled by the man’s reaction, Pritis walked around the desk to see what the man in the lab coat was staring at on the display panel.

At first, he didn’t understand. But as he read the words on the screen, over and over, they first confused, and then alarmed him. The jammer in his hand began vibrating. He looked down at it absently. He had forgotten to turn it off!

A noise outside slipped in through the open door be-hind him. It was the unmistakable roar of jet engines get-ting louder each second with the Doppler resonance of the Blueshift, something hurtling down upon them at incred-ible speed.

In that moment of understanding, Pritis recalled his dream of treasure falling from the sky. Only this time, he saw his father in a flash of gold.

About the Author

Rich Brueckner writes about people and technology at insideHPC. He lives with his young clone in Portland, Oregon.

print 'n fly - sc13

Technology

data intelligence systems

fastest infiniband systems

leading interconnect

explosive growth of

ib fdr technology

anniversary of sc

top500 supercomputers

locals guide