ISC HPC Blog
A Few Words About Obvious Errors
Some of you might have read my last blog and may have wondered why I spoke of T-Systems. Well, it should have been T-Platforms obviously - and it is only my deep involvement with those other German guys over the last 10 years that made me make this mistake. Sorry to T-Platforms – we have changed the blog entry by now.
But then: why not T-Systems? If the future of HPC is in the Cloud then we will need Cloud providers. The Cloud has been one of the hot topics in the last ISC conferences and it figures prominently in the discussions about the program for ISC’13. I remember having been quite outspoken about the Cloud during an Analyst Crossfire session during ISC’11. At that time my prediction was that the Cloud would go away and that HPC should not care too much about it. Was I wrong? In a sense, yes. The Cloud is not about to go away in the near future and having discussed HPC concepts with a number of media companies recently I figured that for them the Cloud is a real option. Now I might save my day by claiming that high throughput computing is anyway not to be mixed up with HPC but that would be too simple.
We see an increasing number of fields that are not able to master the million-core challenge but still need a huge amount of CPU time. Some of these fields are relatively new to HPC – like the media interest in computing or the financial markets with their frenzy for speed in trading non-existing goods for non-existing money.
Just as a side remark: One of my nightmares is to wake up in a world where a Cray system erroneously crashed the New York Stock Exchange while an IBM system by accident sent the Frankfurt Stock Exchange into disaster – both of them working at the speed of light to make sure no human being can ever intervene. These are the days when I start counting my Yen and Yuan and hope that the Japanese and Chinese still refuse to use Western software – hoping that different teams of software developers make different mistakes.
But many of the “farming” applications that we see working in hundreds of copies on hundreds of cores each are good old friends. There are the crash people exploring options and variations. There are our chemistry guys testing hundreds or thousands of combinations. The world is full of well-developed software packages that benefit from the mass of cores available in systems today, but do not necessarily need a tightly integrated system. Scientists and developers are no longer only interested in getting one perfect solution but prefer to have many solutions of many variations to pick the best solution from a sample of results. Genetic algorithms and other optimization methods greatly support such approaches and science and industry both benefit from this.
In such a world the Cloud (or whatever we are going to call this in 10 years from now) is an option for those who need a lot of compute power. And then companies like T-Systems will be among the HPC providers of this world - perhaps using T-Platforms technology.
What all of this will do to traditional HPC centers? Read my next blog.
About the author
Prof. Dr. Dr. h.c. Michael M. Resch is the director of the High Performance Computing Center Stuttgart (HLRS), the director of the Department for High Performance Computing, and the director of the Information Center (IZUS) at the University of Stuttgart/Germany. Michael Resch has over 20 years of experience in HPC and was an invited plenary speaker at SC’07 in Reno/USA. He is a member of the board of the German Gauss Center for Supercomputing (GCS), and is a member of advisory councils for Microsoft, Triangle Venture Capital Group and a number of foundations. Michael Resch is a member of the advisory board of the Paderborn Center for Parallel Computing (PC2) and a member of the steering committee of the IDC HPC User Forum. He has contributed for years to the success of ISC and has been active in a variety of conferences and community activities in supercomputing over the last years. Michael Resch holds a degree in Technical Mathematics from the Technical University of Graz/Austria and a PhD in Engineering from the University of Stuttgart/Germany. In 2002 he held an Assistant Professorship at the department of computer science of the University of Houston/TX. In 2009 he was awarded an honourable doctoral degree from the National Technical University of Donezk/Ukraine. In 2011 he was awarded an honourable doctoral degree from the Russian Academy of Science.
Prof. Dr.-Ing. Dr. h.c. Dr. h.c. Michael Resch