Prof. Kim started researches in the field of neural network implementation from his sabbatical year. For the lightweight neural networks, complex data format and operations are converted into simple data format and operations with negligible performance degradation.

He has interests in applying the approximate computing and stochastic computing to the neural network implementation. Firstly, he developed several iterative structures based on logarithmic multiplications. Secondly, he is studying the application of stochastic processing in the inference system. At this time, it is thought that there are many things to be researched in the training of neural networks with approximate computing and stochastic processing.

Along with the increasing requirements in network security equipment, the industrial needs for the high throughput and multiple pattern matching methods will become enormous in near future. The string matching methods can be applied to the bioinformatics, face recognition, network security, etc. On the other hand, the multiprocessing and hardware acceleration would be prevalent in processing intensive computations and information managements. Due to the slow speed of the software-based string matching in uni-processor environments, the multiprocessing environments will be adopted in order to obtain high-quality string matching engine in many-core era. Therefore, his research will be focused on the high-performance parallel string and pattern matching architecture for multiprocessing environments. In addition, hardware acceleration can adopt bit-parallelism, where several hardware platform can support the hardware acceleration using high speed interface between hardware and host.

For the era of 10G and 100G networking environments, the existing hardware/software techniques cannot guarantee the performance in the networking environments. Both smart offloading technique and novel idea should be considered to support high speed networking environments. In addition, virtual networking environments have been prevalent, which degrades the networking performance. In order to overcome the limitation of the existing hardware/software techniques, algorithms and techniques for enhancing networking performance are being studied.