Low Earth Orbit (LEO) satellite constellation systems are expected to be integrated with terrestrial networks for future communication services over wide areas. For real-time applications, direct data relay between users is effective to minimize the propagation delay via LEO satellites, where multi-beam control is an essential function in LEO satellites to efficiently allocate beam resources over the areas. However, no previous works considered the satellite’s buffer limitations and the differences in throughput between uplink (UL) and downlink (DL) which caused queuing delays. This paper proposes a coordinated UL/DL multi-beam scheduling to minimize the end-to-end latency of data relay by considering the available buffer space of the satellite and variations in UL/DL throughput due to satellite movement. An optimization problem is formulated for the scheduling design to minimize the end-to-end latency. We demonstrate with simulation results that the proposed scheduling method reduces the latency by 16.0%, compared to the conventional methods.

In this paper, we propose a novel resource management technique for heterogeneous supercomputing systems affected by manufacturing variability. Our proposed technique called VAHRM (Variation-Aware Heterogeneous Resource Management) takes a holistic approach to job scheduling on highly heterogeneous computing resources. VAHRM preferentially allocates energy-efficient computing resources to an energy-consuming job in a job queue, considering the impact on both the job turnaround time and the power consumption of individual resources. Furthermore, we have developed a novel approach to modeling the power consumption of computing resources that have manufacturing variability. Our approach called TSMVA (Two-Stage Modeling with Variation Awareness) enables us to generate the first variation-aware GPU power models, which can correctly estimate the power consumption of each GPU for a given job. Our experimental results show that, compared to conventional first-come-first-serve (FCFS) and state-of-the-art variation-aware scheduling algorithms, VAHRM can achieve respective improvements in system energy efficiency of up to 5.8% and 5.4% (4.5% and 4.2% on average) while reducing the average turnaround time of 21.2% and 11.9%, respectively, for various workloads obtained from a production system.

Polar codes are provably capacity-achieving errorcorrecting codes suitable for error correction and source coding, constrained coding, and multiple access channels. Many decoding schemes have been proposed for Polar codes, including CRCaided successive cancellation list decoding (CA-SCL), known for its superior error-correcting performance. Belief propagation (BP) techniques have also been explored for decoding polar codes. BP decoding with partially permuted factor graphs (PPFG) exhibits good performance despite being inferior to CA-SCL decoding. Guess Random Additive Noise Decoding (GRAND) is a general decoding scheme for linear codes. Soft-GRAND (SGRAND) exhibits excellent performance for Polar codes despite its high time complexity. We introduce a novel decoding scheme that combines BP decoding with PPFG and SGRAND, advancing the state-of-the-art in Polar code decoding. Our proposed scheme, which fuses BP decoding with PPFG and SGRAND, demonstrates performance similar to that of CA-SCL decoding, representing a significant step forward in Polar code decoding.

CUDA toolkits are widely used to develop applications running on NVIDIA GPUs. They include compilers and are frequently updated to integrate state-of-the-art compilation techniques. Hence, many HPC users believe that the latest CUDA toolkit will improve application performance; however, considering results from CPU compilers, there are cases where this is not true. In this paper, we thoroughly evaluate the impact of CUDA toolkit version on the performance, power consumption, and energy consumption of GPU applications with three GPU architectures. Our results show that though the latest CUDA toolkit obtains the best performance, power consumption, and energy consumption for many applications in most cases, but we found a few exceptions. For such applications, we conducted an in-depth analysis using the SASS to identify why older CUDA toolkit achieve performance improvement. Our analysis showed that the factors that caused them are by three phenomena: aggressive loop unrolling, inefficient instruction scheduling, and the impact of host compilers.

Understanding the variations in performance and power-efficiency of compute nodes is important for enhancing these factors in modern supercomputing systems. Previous studies have focused on variations in CPUs and DRAMs, but there has been little attention on GPUs. This is despite many current supercomputing systems employing GPUs (which consume a significant fraction of the power of such systems) as power-efficient accelerators for HPC applications. This paper describes the first thorough evaluation of performance and power-efficiency variations in GPUs. Specifically, we execute 48 CUDA kernels on 856 devices selected from three generations of NVIDIA GPUs (P100, V100, and A100), and analyze the impact of differences in both the CUDA kernels and GPU generation on performance and power-efficiency. Our analysis shows that there are non-negligible variations in both performance and power-efficiency, and that these variations are strongly affected by both the kernels that are running and the GPU generation.

We developed soccer ball-shaped interactive artifacts (COLUMN) consisting of eight modules that are connected to twelve servomotors. In the interaction, COLUMN becomes a social mediator to prompt the connectivity of the users. We claim that the above process generates a variety of user's interpersonal coordinates and these can be mapped to the shape of COLUMN's. Therefore, the motivation of our study is to extract the patterns of user's interpersonal coordination. It is also interesting to explore how these patterns lead to imagine behavior for an artifact which has different degree of freedom of its body than the existing.

In this paper, we propose control system which supports unicycle operation for human's skill up. Therefore, the unicycle operation analysis method is proposed first, and it is shown usefulness by experiment. The element used for analysis are rider's CP (Center of Pressure) and MBP (Moment Balancing Point) which negates a saddle rotation moment. Using those elements, we propose evaluation method of unicycle rider's operation. Dynamic unicycle simulator is used for supports unicycle operation for human's skill up.