.
Distributed AI systems face critical memory management challenges across computation, communication, and deployment layers. RRAM based in memory computing suffers from scalability limitations due to device non idealities and fixed array sizes. Decentralized AI frameworks struggle with memory efficiency across NAT constrained networks due to static routing that ignores computational load. Multi agent deployment systems tightly couple application logic with execution environments, preventing adaptive memory optimization. These challenges stem from a fundamental lack of coordinated memory management across architectural layers. We introduce Self Evolving Distributed Memory Architecture for Scalable AI Systems, a three layer framework that unifies memory management across computation, communication, and deployment. Our approach features (1) memory guided matrix processing with dynamic partitioning based on device characteristics, (2) memory aware peer selection considering network topology and computational capacity, and (3) runtime adaptive deployment optimization through continuous reconfiguration. The framework maintains dual memory systems tracking both long term performance patterns and short term workload statistics. Experiments on COCO 2017, ImageNet, and SQuAD show that our method achieves 87.3 percent memory utilization efficiency and 142.5 operations per second compared to Ray Distributed at 72.1 percent and 98.7 operations per second, while reducing communication latency by 30.2 percent to 171.2 milliseconds and improving resource utilization to 82.7 percent. Our contributions include coordinated memory management across three architectural layers, workload adaptive resource allocation, and a dual memory architecture enabling dynamic system optimization.
This study investigated the motion characteristics of ball-trapping techniques in soccer, focusing on directional changes in spaces between the defense and midfield lines by comparing advanced and intermediate university-level players. Effective ball-trapping techniques are critical for maintaining stability during play and for enabling offensive strategies. Although limited studies have been conducted on ball trapping, often using oversimplified and closed-skill environments, this study identified the technical factors that enhance these skills to improve performance in match settings. Motion-capture technology focuses on ball-to-foot contact control, pelvic orientation, and ankle-eversion angular speed. Ten advanced and eight intermediate soccer players participated in this study, and each performed multiple directional ball-trapping trials under standardized conditions. The advanced group demonstrated significantly shorter total trapping times than the intermediate group ( p < 0.05). The vertical distance between the center of the ball and the ball-to-foot contact point was larger and the horizontal distance was smaller in the advanced group ( p < 0.05). Additionally, the advanced players exhibited pelvic orientational angles that were more parallel to the ball trajectory and demonstrated higher ankle-eversion angular speeds during impact ( p < 0.05). These results suggest that advanced players minimize trapping errors through precise ball-to-foot contact adjustment, proactive pelvic rotation, and effective utilization of ankle flexibility and strength. These technical elements enable smoother transitions and improved directional changes during play. The findings provide valuable insights for coaching, emphasizing drills to enhance ball-to-foot contact precision, pelvic rotation, and ankle flexibility, improving overall trapping performance.
This study explored the relationship between floating toes and athletic performance among collegiate track and field athletes. A total of 422 athletes (sprinters, jumpers, and distance runners) and 136 controls participated in this study. Plantar surface images were captured using a specially designed foot scanner during standing to calculate the floating toe score. The score, which sums the floating toe points of all toes, categorizes them into ‘floating toe’, ‘incomplete contact’, or ‘normal toe’. The World Athletics score served as a measure of athletic performance. Overall, collegiate track and field athletes had a significantly higher mean floating toe score (14.22 ± 4.87 points) than controls (11.06 ± 6.05 points) (p < 0.05), suggesting that fewer floating toes may confer a performance advantage. Within the track and field events, sprinters and jumpers showed significantly higher floating toe scores than distance runners (p < 0.05), emphasizing the variations in floating toes across different events. Although no significant correlation between floating toe score and World Athletics score was observed overall, a significantly negative correlation was found among distance runners (r = -0.25, p = 0.006), indicating that distance runners with higher athletic performance tend to have more pronounced floating toes. These findings suggest a nuanced relationship between floating toes and various aspects of track and field performance, particularly in distance running.
PurposeTo compare screw surface characteristics between hemi-engaging and non-engaging implant-supported fixed partial denture (FPD) designs after cyclic loading. Materials and MethodsTwenty-four implants measuring 4.3 x 10 mm were mounted on acrylic resin blocks. Specimens were divided into two groups. An experimental group included twelve 3-unit FPD with a hemi-engaging design; a control group included twelve 3-unit FPDs with the conventional design of two non-engaging abutments. Both groups were subjected to two types of cycling loading (CL), first axial loading, and then lateral loading at 30 degrees. Load was applied to the units one million times (1.0 x 10(6) cycles) for each loading axis. Data on screw surface roughness in three locations and screw thread depth were collected before (BL) and after (AL) each loading type. Screw surface roughness was measured in mu m using a mechanical digital surface profilometer and optical profiler. To evaluate screw thread depth in mu m, an upright optical microscope Axio-imager 2 was used. To confirm readings made from the optical microscope, four random samples were selected from each group for scanning electron microscopy (SEM) analysis. The effect of cyclic loading was evaluated by averaging values across the two screws within each specimen, then calculating difference scores (DL) between BL and AL (DL = AL - BL). Additional difference scores were computed between the non-engaging screws in each experimental group specimen, and one randomly selected non-engaging screw in each control specimen. This difference was referred to as the non-engaging DL. Statistical significance was assessed using Mann-Whitney U tests (alpha = 0.05). ResultsComparisons of DL and non-engaging DL by loading type revealed one significant difference regarding surface roughness at the screw thread. Significantly greater mean changes were observed after axial loading compared to lateral loading regarding both DL (axial M = -0.36 +/- 0.08; lateral M = -0.21 +/- 0.09; U = 20; p = 0.003) and non-engaging DL (axial M = -0.40 +/- 0.22; lateral M = -0.21 +/- 0.11; U = 29; p = 0.013). No significant differences in screw surface roughness in other sites or thread depth were found between the experimental and control abutment designs in DL or in non-engaging DL. No significant differences were found for DL (axial U = 13, p = 0.423; lateral U = 9, p = 0.150;) or non-engaging DL (axial U = 13, p = 0.423; lateral U = 18, p = 1.00). ConclusionsResults suggest that overall, changes in screw surface physical characteristics did not differ between hemi-engaging and non-engaging designs after evaluating screw surface roughness and thread depth before and after axial and lateral cyclic loading.
Education about menstruation is a sensitive topic for young female students; providing appropriate knowledge is essential for maintaining and improving their health. The present study was conducted to collect data corresponding to different factors affecting health among young individuals; the menstrual status, exercise habits, sleep status, and body composition of these individuals as well as the relationships among these factors, were evaluated. Altogether, 200 female students responded to the survey; 129 completed all the physical measurement items. As a case study, face-to-face interviews regarding menstrual symptoms were conducted. Results showed that 49/200 (25%) and 120/200 (60%) participants experienced moderate or severe pain before and during menstruation, respectively. The degree of pain one week before menstruation and during menstruation were significantly positively correlated (r = 0.573, p < 0.01). When analyzed as group data, it was difficult to identify the relationship between menstrual status, exercise habits, and sleep status; these were found to be intricately associated with various factors. The case study analysis confirmed that some individuals experienced physical and psychological symptoms, such as irregular menstrual cycles, premenstrual syndrome, and severe menstrual cramps.
