The Jetson Charger Cluster is a cluster from 25 Jetson TK1 platforms, which are bound together using a high-speed gigabit network. 24 working node are visible on the front of the JCC, while the master node is hidden behind together with network and power interconnects. JCC can be used for solving real world problems, such as weather forecasting, climate modelling, computer vision, and space sciences, just to name a few. At the same time, JCC provides energy efficiency, and low cost which is equivalent to that of a single high end workstation.
Poster: [JCC]
JETSON TK1 DEV BOARD
Each Jetson TK1 Development Platform is designed by NVIDIA. It includes a Tegra K1 system-on-a-chip (SOC), 2 GB memory, 16 GB eMMC memory, and a number of input/output interfaces.
NVIDIA TEGRA TK1 SoC
Each Tegra K1 system-on-a-chip includes a quad core ARM Cortex A15 Processor that can run at maximum clock rate of 2.3 GHz. It also contains a smaller A7 core and 192 NVIDIA CUDA cores. In total, this is 100 ARM Cortex processors cores, 25 ARM A7 processor cores, and 4,800 CUDA cores; 50 GB of RAM memory, and 400 GB of flash memory.
CREDITS
The JCC was conceived by the UAH Computer Engineering faculty. Mr. Jason Winningham designed the JCC’s enclosure, power distribution, and interconnection network. Mr. Armen Dzhagaryan contributed to the design, software installation, and initial testing. The financial support is provided by the ECE Dept. Interim Chair, Dr. Earl Wells, and by an NVIDIA’s donation. Contact: Dr. Aleksandar Milenković, This email address is being protected from spambots. You need JavaScript enabled to view it. .