IBM today announced the availability of its world renowned Blue Gene supercomputing system at its newest Deep Computing Capacity on Demand Center in Rochester, Minn.
The new Center will allow customers and partners, for the first time ever, to remotely access the Blue Gene system through a virtual private network and pay only for the amount of capacity reserved.
Blue Gene provides customers supercomputing speed at a peak performance of 5.7 teraflops, with a single full-rack system.
“IBM has been reaching out and working with a number of our key business partners to port applications to build the Blue Gene ecosystem,” said David Gelardi, vice president, IBM Deep Computing Capacity on Demand. “Customers have been able to access IBM’s Deep Computing solutions via our centers for more than 18 months.”
Expanding the Gene Pool
Gelardi said today’s announcement provides customers with another way to “test drive” Blue Gene to help them decide whether to purchase their own racks or continue to buy time.
While some of the most ambitious supercomputing work still takes place in government labs, IBM said the potential for Deep Computing breakthroughs in new commercial markets — such as drug discovery, product design, simulation, animation, and financial and weather modeling — is growing rapidly.
Illuminata analyst Jonathan Eunice told TechNewsWorld that only a small number of people will touch Blue Gene. But, he added, the value to those who do is very high.
“The historical problem with this type of equipment is that it is typically inaccessible,” Eunice said. “You can read about it and read the architecture, but you have no way of getting a hold of it as a generalist. It’s not that every Windows or Linux programmer is going to want access, but this broadens the number of people who can get access and do some interesting things.”
IBM, working with its business partners, is making Blue Gene applicable for workloads across a variety of disciplines. Many national lab and university members are enabling a growing list of HPC applications in areas of life sciences, hydrodynamics, quantum chemistry, molecular dynamics, astronomy and space research, and climate modeling.
Other areas of interest include financial modeling, business intelligence, risk and compliance, aerodynamics study and testing and manufacturing processes.
Many key software vendors such as Novell-SuSE, LSTC and Allinea have shown an interest in enabling their applications on Blue Gene. Etnus is already in the process of enabling their TotalView parallel debugger to Blue Gene.
Eunice said broadening access to Blue Gene is a good move for Big Blue, considering that parallel systems providers have had difficulty finding enough skilled people in the world to program the machines.
“IBM has solved some of the skills problem by basing Blue Gene on Power architecture,” Eunice said. “There are a lot of tools and compilers who understand Power assembly language and can optimize for Power.”