Intel has uncovered a machine programming research system that helps autonomously identify errors in code.
During the ‘Intel Labs Day’ digital event company mentioned that preliminary tests of the system called ControlFlag trained and learned novel defects on over one Billion unlabeled lines of production-quality code. According to a study, software developers spend nearly 50 percent of their time debugging. Intel is an American multinational firm and technology company as well as it is the world’s largest and highest valued semiconductor chip manufacturer.
Justin Gottschlich, principal scientist and founder of Machine Programming Research at Intel Labs commented that with the help of ControlFlag software developers can spend notably less time debugging, so that human software programmers do best expressing creative, new ideas to machines. It cost $1.25 trillion every year as software development costs and 50 % is spent on debugging code. The reason for taking machine programming is that there is an emergence of heterogeneous hardware, that it no longer was going to be just the CPU, it has now evolved into FPGA, GPU, ASICS, and so on.
As stated by Intel, it is difficult to find a software developer who has the expertise to correctly, efficiently, and securely program across diverse hardware that introduces another opportunity for new and harder-to-spot errors in code. Intel company mentioned that ControlFlag could help this challenge by automating the tedious parts of software development, such as testing, monitoring, and debugging.
The bug detection abilities of ControlFlag are enabled by machine programming. Machine programming is a combination of programming languages, machine learning, formal methods, computer systems, and compilers. It is a self-monitored system that operates through a capability known as anomaly detection. It learns from verified models to identify normal coding patterns, identifying anomalies in code that are expected to cause a bug. ControlFlag can detect these anomalies despite any programming language.
Intel has started to use ControlFlag internally to identify bugs in its software and firmware product improvement.
