Background
With the rapid development of mobile internet applications and the expansion of cloud computing and big data, the data center has updated from the traditional mainframes and minicomputers to the x86 server. Since people have a new need for the pattern of server workloads, more and more intelligent scenes are deployed by small servers.
Introduction
The solution of cluster servers in the combination mode of multiple core-boards provides standard software and hardware interfaces. It supports distributed AI computing which can be applied to machine learning, artificial intelligence and big data development. Compared with the traditional cluster servers, it is characterized by low power consumption, high manageability and great flexibility.
And the solution supports the simultaneous operation of multiple core-boards like Core-3399Pro-JD4, Core-3399-JD4 and Core-3328-JD4. The maximum number of core-boards it supports can be up to 11 at one time. And it also runs the Android and Linux systems steadily in 7x24 hours, achieving unified debugging and firmware burning.
Construction
Combination of Cluster Server Backplane and Core Board.
Features
Simultaneous operation of multiple core-boards
It supports multiple core-boards to work at the same time or their respective operation with more flexible applications achieved in the project.
Flexible choices of core-board types
It can adjust the computing power and costs with any number or different core-boards.
AI accelerated processor NPU
The core-boards possess the function of AI acceleration with NPU, and can improve the AI computing power for several times through the combination of every NPU.
Multiple operating systems supported It is able to run the Android and Linux operating systems in a stable and reliable way.
It can flexibly deploy several chatting tools to organize a marketing system and it's equivalent to 11 computers running at the same time.
Debug serial port × 11, Power input × 1, USB2.0 × 2, RJ45 Gigabit Ethernet × 4, Work Indicator × 11, Network connection indicator × 11, Fan power × 2, Burning debugging serial port × 1, Reset × 11, Recovery × 11, Custom LED × 11
It is widely used in machine learning, high-performance computing/storage, computer vision, edge computing, big data development, game distribution and other scenarios.
To know more about the solution, please send an email to sales@t-firefly.com
| 
