First of all, the European attendance to SC remains steady and meaningful. This despite the distance, the worsening service on the airlines, the shrinking space on board for your arms and legs - flying over the atalantic costs less but it is more and more difficult to find suitable connections, let alone a seat - the growing costs associated with lodging in a reasonably comfortable hotel, and the time of the year when the event takes place, usually November which, except for Florida, Arizona and California, is not the best one in the USA from a meteorological point of view.
Why do we Europeans, year after year, enthusiastically respond to SC call and rush there by the tens - if not hundreds? Reasons are many: let me try to identify a few, not necessarily scientific ones.
1. We don't have, despite many efforts, any such event in Europe. Cherry Pancake said in her welcome address ? SC is a unique gatheriing. True, HPCN Europe, though a remarkable initiative , has never represented for the HPCN users and vendors community the way SC does. Limited and not fully representative attendance, academic predominance in designing and implementing the program, high costs for the show space (some of the vendors felt that they had to participate in order not to displease the European Commission which has been always fully supportive of the Initiative).
2 - The HPCC global market is highly dynamic.Whether it has to do with technological and architectural advances, operating systems ,new products, new companies growing,old companies shrinking or dying or reviving etc.etc. there is always something new ,and there is always the expectation that secrets will be uncovered at SC and the gospel will be revealed to the lucky attendees. They will be the first ones to know, then , and coming back from the USA therefore they will be more knowledgeable than those who decided to stay in Europe and do not travel.
3. To see and to be seen: some may disagree, but it is my impression that being seen at SC is felt as it would add something to ( or confirm ) one?' professional involvement with the HPCC world .If you are not there, it?' like "missing a cycle " and certainly (remember that it is a unique gathering ) chances of seeing other professionals who deal with the same business,or discipline or area of interest worldwide are very high.If you are an independent consultant,for example, this alone repays a great part of the costs incurred.
If I may expand more on this issue, it is not unusual that people who live and work in Europe almost "next door" to each other, at a given time of the year, defer a potential business meeting to when we both are at SC: which does not always prove to be feasible since,once you are there, the schedule is generally speaking very tight, between presentations, visits to the exhibitors, unscheduled meetings, reunions, parties, dinners etc.
4. Reward : in many cases, specially in the Academic field, a visit to SC is used as a reward to those cooperators in the HPCC field who have well deserved in the past twelve months.And being the month of November close to the year-end, the remains of the budget (or budget money found somewhere else through administrative wizardies ) are being used for this purpose .Money well spent,in my opinion,as it will certainly increase the professional knowledge of the individuals,let alone the motivation.
When it comes to the Conference, finally, it was, as usual, well attended : the technical program,t hough, didn?' include much novelties Also, since quite a few years, the key-note speech does not have in my opinion (and with all the respect for the speakers ) that touch of global scientific and professional authority that one would expect in a global event like this one. I personally miss those years when the Conference Key-noter gave a sort of "Supercomputing at a glance" speech, which was regarded as a "gospel" for the coming years .
The show was very well run, and the decreasing number of large HPCC Vendors was compensated by a wealth of other large and small technologies and services manufacturers and providers,which,in one way or another, increasingly integrate into the HPCC world (a record of 157 Exhibitors). Hewlett-Packard, SGI , IBM, Compaq (with proper visibility given to the technological partnership with the European QSW) SUN and,once again the Japanese (NEC, HITACHI and FUJITSU ) had, as usual , the best spots of the show.TERA was there too : in my opinion,however, now that some of the most critical technical goals have been achieved, the Company must show more substantial commercial results to gain users confidence.Among the Japanese Vendors, NEC was the one which showed the best mix of hardware / technology advances and market results (mainly from Europe and Japan, though, since the US market is still sort of "off limits" for them because of the well known importation tariffs issue).
Many expected, during the show, an announcement regarding the future of Cray Research (as an old-timer, and having grown up professionally in the Seymour Cray years, I must confess that I was among those many).Which didn't come,unfortunately : the clouds were still there,and the Cray booth didn?t have much to say,let alone to show.
The fast progress of the Industry was felt everywhere: just to name a couple of facts, one might be interested to know that in 1993 the fastest computer on the floor was capable of a peak performance of 100 Gflop/s. Such performance was available this year on a desk-top? Similarly, the SCInet 1993 was capable of 155 Mbit per second, versus the 10 Gbit/s of this year.
All in all, it was worth attending: see everybody in Dallas, Texas, next year!