Junsu Kim

I am an M.S. student in the Department of Electrical Engineering at Korea University, under the guidance of Prof. Yunho Oh.

My research interests lie in the areas of Computer Architecture and Systems, Memory Systems, and Artificial Intelligence (AI). I aim to bridge the gap between computer architecture and emerging applications via algorithm-hardware co-design. Recently, I have focused on orchestrating heterogeneity in computer architecture to achieve high performance and energy-efficiency for emerging applications. My research contributions include:

Heterogenous Memory: Developed a memory oversubscription-aware tensor migration scheduling for a GPU unified storage architecture [P1]. Enabled unified Flash Translation Layer (FTL) with dedicated IOMMUs and the optimized migration scheduler for hetergeneous memory-based DNN accelerator systems [P2]. Analyzed the behavior of a real CXL-based system on datacenter workloads to uncover open challenges in the emerging hardware platform [P4].
Heterogenous Processors: Enabled software support for arbitrary integer formats on GPUs by software-assisted packing and load balancing between Tensor, INT and FP CUDA cores [ICPP’24]. Accelerated Yinyang K-Means via dynamic thread allocation for CPU and GPU on mobile platform [P5].
High Performance AI Frameworks: Developed dynamic thread allocation for a multi-tenant GPU [ESL’24]. Enabled hardware supports for bitslice representation to process arbitrary low-precision numeric formats on GPUs [P3]. Developed memory-efficient continual learning [AAAI’25, P6].

Before I joined Korea University, I worked on developing memory-efficient continual learning frameworks with Dr. Suhyun Kim at Korea Institute of Science and Technology (KIST) in 2022 [AAAI’25, P6] and on Rowhammer mitigation with low overhead at Hanyang University in 2021 [ICCD’21]. I received my B.S. with honors from Hanyang University in 2021.

selected publications

PhysRev

Can Quantum-Mechanical Description of Physical Reality Be Considered Complete?

A. Einstein^*†, B. Podolsky^*, and N. Rosen^*

Phys. Rev., New Jersey. More Information can be found here , May 1935

Abs DOI HTML PDF Video

In a complete theory there is an element corresponding to each element of reality. A sufficient condition for the reality of a physical quantity is the possibility of predicting it with certainty, without disturbing the system. In quantum mechanics in the case of two physical quantities described by non-commuting operators, the knowledge of one precludes the knowledge of the other. Then either (1) the description of reality given by the wave function in quantum mechanics is not complete or (2) these two quantities cannot have simultaneous reality. Consideration of the problem of making predictions concerning a system on the basis of measurements made on another system that had previously interacted with it leads to the result that if (1) is false then (2) is also false. One is thus led to conclude that the description of reality as given by a wave function is not complete.