The TOP500 list clearly demonstrates the dominant position the U.S. assumes in the world both as producer and as consumer of high performance computers. In Table 3 the total number of installed systems in the major world regions is given with respect to the origin of the computers.
If one considers in Table 3 the country of origin then it is striking that 418 out of the TOP500 systems are produced in the U.S., which amounts to 84% of all installed systems. Japan accounts for 14% of the systems, and Europe produces only 2.4%. The extent of the American dominance of the market is quite surprising, and has been even increasing from the previous year, when the U.S. share was 78%. For years, in particular in the mid 80's, there were ominous and ubiquitous warnings that the American supercomputer industry (which was essentially Cray Research at that time) is highly vulnerable to an ``attack" by the Japanese vertically integrated computer giants Fujitsu, NEC, and Hitachi. Obviously this has not happened. How much various efforts such as the NSF Supercomputing Initiative in the mid 80's, or more recently the HPCC Program have contributed to the current vast superiority of the U.S. high performance computing industry, remains to be investigated. It is interesting to note that one view expressed outside the U.S. [13] is that strengthening the U.S. HPC industry and easing the transition to MPP was the only rationale for the HPCC Program.
The numbers for Europe are actually better than last year (12 machines in 1994 versus 7 machines in 1993). However, in last year's report the ``European" machines included Fujitsu VP products, which are resold in Europe by Siemens/Nixdorf. In spite of a recent installation of a European system at a U.S. government laboratory (a Meiko CS-2 at Lawrence Livermore Natl. Lab. ) the situation in Europe remains bleak. With lack of immediate access to the newest hardware, and the absence of the close interaction of users with vendors as is prevalent in the U.S., the best the European High Performance Computing and Networking Initiative can accomplish is maintaining the status quo of Europe as a distant third in high performance computing technologies.
Table 3: US Share of Total Number of Installed TOP500
Systems
Table 4 is analogous to Table 3, but instead of the number of systems, the aggregate performance in Rmax-Gflop/s is listed. Table 5 lists the ratio of the corresponding entries in Tables 3 and 4, i.e., the average Rmax-Gflop/s per machine. From Table 5 we can see that on average the machines manufactured in Japan have higher performance ratings than machine manufactured in the U.S. Also machines installed both in Japan and the U.S. appear to be more powerful than the machines installed in Europe or in other countries.
Table 4: US Share of Total Rmax (in Gflop/s) of Installed
TOP500
Systems.
Table 5: Average Gflop/s per machine.
A more interesting analysis of Tables 3 and 4 addresses the ongoing question of who has the higher trade barriers with respect to high performance computing, the U.S. or Japan? Table 3 shows that obviously both countries favor their own machines over their competitor's. But let us assume for the sake or argument that the European market is equally open and accessible to both American and Japanese machines, i.e., the distribution of supercomputers in Europe reflects an open market. According to the distribution in Europe the split in an open market should be 93% American and 7% Japanese. Assuming these market shares worldwide, the number of Japanese machines in the U.S. would be 17 (as opposed to 5), and the number of U.S. made supercomputers installed in Japan would be 76 (as opposed to 30). Clearly both countries seem to have mechanisms in place which restrict competition in favor of the local products. The reader can decide whether only 5 out of 17 or only 30 out of 76 constitutes a more biased situation.