Electro-Optical Sciences (EOS) will be the first life sciences company to use on-demand supercomputing power from IBM to analyse data collected by a handheld skin-scanning device for detecting malignant melanoma.
By submitting your personal information, you agree that TechTarget and its partners may contact you regarding relevant content, products and special offers.
IBM's Deep Computing on demand initiative allows customers to access supercomputers via the internet whenever the customer needs that level of computing power.
The idea is to encourage development of applications and products that require supercomputing technology without the requirement that expensive supercomputers be bought and maintained.
"When they need it, we will supply it," David Jursik, vice-president of sales for IBM Deep Computing said.
The centre, which has Linux and Unix clusters, opened in June and had attracted customers in the oil industry, although IBM said that it expected computing on demand would be attractive to life sciences customers.
The EOS product, called MelaFind, uses statistical pattern recognition to analyse a database containing more than 20,000 high-resolution images of skin lesions to compare with images taken during skin scans at doctor's surgeries, said EOS.
Although melanoma accounts for just 4% of skin cancer cases, it causes 79% of skin cancer deaths and early detection of the disease is critical.
EOS has developed its device for early detection of malignant melanoma, but needed a way to analyse data obtained from the scans quickly and efficiently.
Clinical tests indicate that MelaFind's computational method is more effective than experienced dermatologists in detecing melanoma early.
The scanning takes about a minute and then data from the handheld device is sent to IBM's supercomputing centre where an analytical engine compares images with those stored in a database. Within five or 10 minutes, the comparison of images is finished, Jursik said.
IBM expected companies in other industries to become computing-on-demand customers, Jursik said. IBM targeted industries with "obvious" supercomputing needs, but is hearing from companies in, for instance, online gaming and movies.
IBM also will announce that Paradigm Geotechnology, which provides petroleum geoscience and drilling technology, has bought three IBM supercomputers to help create images of deep underground oil reservoirs that are more accurate than existing models.
The supercomptuers are made up of several hundred IBM Blade Centre nodes installed in Paradigm's offices in Texas, Woking, UK, Moscow and Mumbai.
Nancy Weil writes for IDG News Service