IBM Israel Science and Technology Ltd IBM has the world’s largest IT research organization, with more than 3,000 scientists and engineers working at 8 labs in 6 countries. IBM invests more than $5 billion a year in R&D and is the world’s leader in patent filings.
In aggregate, the company holds nearly 37,000 patents worldwide. IBM Israel Science and Technology Limited is better known as IBM Research – Haifa. Since it first opened as the IBM Scientific Center in 1972, the Haifa lab has conducted decades of research that have proved vital to IBM’s success. The lab is one of five research laboratories located outside of the United States, and has close working relationships with IBM Israel and its twin research laboratory in Zurich. In Haifa, 25 percent of the technical staff has doctorate degrees in computer science, electrical engineering, mathematics, or related fields. Employees are actively involved in teaching in Israeli higher education institutions and in supervising post-graduate theses. R&D projects are being executed today in areas such as storage systems, cloud computing, healthcare and life sciences, verification technologies, business transformation, information retrieval, programming environments, optimization technologies, and analytics.
In code optimization technologies, HRL focuses on a range of optimization problems, both in compilers and post-link optimization tools which take advantage of the underlying structure of modern VLIW and superscalar processors (e.g. PowerPC). Haifa’s researchers contributed to IBM’s proprietary technology components as well as to open source tools such as the GNU C Compiler (GCC). In the context of GCC, IBM Haifa has been involved in researching sophisticated optimization techniques such as global and local instruction scheduling, modulo scheduling, auto-vectorization, auto-parallelism etc. Complementary to optimizing compilers, HRL developed post-link binary instrumentation and optimization technology, called FDPR-Pro. The FDPR-Pro tool uses the profiling information and the program global view to optimize the most frequently executed pieces of code, potentially at the expense of the rest of the program. This technology proved itself in a variety of market segments such as database engines, embedded software components, etc. The Haifa researchers started looking into new promising domains such as dynamic and adaptive monitoring and optimization of long running programs, architectures of media processors (e.g. DSPs) and related code optimization problems, exploitation of machine learning techniques for iterative optimization, etc.