IBM announces the public availability of Milepost GCC, an open-source machine learning compiler that IBM says allows applications to be developed, tested and optimized 10 times faster than current tools.IBM announced June 3 the public availability
of Milepost GCC, a first-of-its-kind open-source machine learning compiler.
The Milepost GCC compiler "intelligently optimizes applications,
translating directly into shorter software development times and bigger
performance gains," IBM officials said
in a news release.
IBM's Research Lab in Haifa,
Israel, worked with
academia and private industry in the European Union to polish the new compiler,
which, unlike commercially available compilers, employs artificial intelligence
to tweak individual pieces of code.
With the compiler, applications can be developed, tested and optimized 10 times
faster than current tools, according to IBM.
Moreover, performance of the software programs can be improved by an average of
18 percent, IBM officials said. These
improvements are significant, given that the average company "devotes 30
to 50 percent of its entire technology infrastructure to the development and
testing of software," according to IBM.
IBM reported that it experienced the 18
percent performance improvement on embedded-application benchmarks conducted on
IBM System p servers. The new compiler is a
result of collaboration between IBM and its
partners in the European Union-funded Milepost consortium.
IBM officials said the compiler is
expected to "reduce time-to-market for new software designs. ... For
example, when a company wants to develop a new mobile phone, it normally takes
application developers many months to get their software running at an
acceptable level of performance. Milepost GCC can reduce the amount of time it
takes to reach that level by a factor of 10."
"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," Bilha Mendelson, manager of
code optimization technologies at IBM
Research Haifa, said in the release. "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."
"We've developed a more cost-effective development process where you
can choose to integrate additional functionality or use less power in your
current system," Milepost Project Coordinator Mike O'Boyle, professor of
computer science at the University of Edinburgh's School of Informatics, said
in the release. "Previously, the same devices could only support a limited
list of features while still maintaining a high level of performance.
Significantly boosting an application's performance means there's now more room
for added functionality while maintaining high performance."
According to the release:
As a by-product of the Milepost
technology, the consortium has launched a code-tuning web site available to the development community. Developers can upload their software
code to the site and automatically get input on how to tune their code so it
works faster.
The Milepost GCC compiler is available
to everyone as of June 25 from the consortium's website http://www.milepost.eu.
The project consortium includes the IBM Haifa
Research Lab, Israel; the University of Edinburgh, UK; ARC International Ltd.,
UK; CAPS Enterprise, France; and INRIA, France.