Since 2010, Haidar held a research scientist position at the Innovative Computing Laboratory at the University of Tennessee. Before that, he conducted research at the INPT University (National Polytechnic Institut of Toulouse). There, he worked at the CERFACS Lab in the Parallel Algorithms Team. His research revolves around Parallel Linear Algebra for Scalable Distributed Heterogeneous Architectures such as multicore CPUs and accelerators (Intel Xeon-Phi, NVIDIA and AMD GPUs). His goal is to create software that simplifies development of applications that achieve high-performance and portability. Such programming models rely on asynchronous and out-of-order scheduling of operations. These concepts are the basis for scalable and efficient software for Computational Linear Algebra and applications. Another research interest is the development/implementation of numerical algorithms and software for large scale parallel sparse problems in order to develop hybrid approaches combining direct and iterative algorithms to solve systems of linear algebraic equations with large sparse matrices.