IBM offers open source machine learning compiler
Subscribe now for $100 (23 issues) and save more than 37% off the cover price!
Get the latest news from Computerworld delivered via email.
Sign up now
IBM has announced the availability of an open source machine learning compiler, which the company says intelligently optimises applications, meaning shorter development times and bigger performance gains.
Called Milepost GCC, the compiler is the result of a collaboration between IBM and partners in the European Union-funded Milepost consortium. The project is an extension of the GCC (GNU Compiler Collection) effort.
IBM expects the compiler to dramatically reduce time-to-market for new software designs. Applications can be more quickly tuned for a targeted architecture. Initial IBM experiments on IBM System P servers had an average of an 18 percent performance improvement on embedded application benchmarks, IBM says.
"Our technology automatically learns how to get the best performance from the hardware — whether mobile phones, desktops, or entire systems — the software will run faster and use less energy," said Bilha Mendelson, manager of code optimisation technologies at IBM Research. "We opened the compiler environment so it can access artificial intelligence and machine learning guidance to automatically determine exactly what specific optimizations should be used and when to apply them to ramp up performance."
The Milepost GCC compiler is available from the consortium website. Also, the consortium has launched a code-tuning website available to developers.