公开(公告)号：US20210004666A1

公开(公告)日：2021-01-07

申请号：US16894276

申请日：2020-06-05

Applicant: ELECTRONICS AND TELECOMMUNICATIONS RESEARCH INSTITUTE

Inventor： Yong-Joo KIM ,  Tae-Ho KIM ,  Eun-Ji PAK ,  Young-Mok HA

IPC: G06N3/063 ,  G06N3/08

Abstract: Disclosed herein are a neural network generation method for neuromorphic computing and an apparatus for the same. The neural network generation method for neuromorphic computing includes selecting at least one part of a brain corresponding to a neural network function requested to be generated by an application, determining whether an existing neural network corresponding to the at least one part of the brain is present in a neural network database, when it is determined that no existing neural network is present in the neural network database, generating new neural network configuration information corresponding to the part of the brain based on a brain information database, generating a new neural network by mapping the new neural network configuration information to neuromorphic hardware based on the new neural network configuration information, and storing the new neural network in the neural network database.

公开(公告)号：US20240127034A1

公开(公告)日：2024-04-18

申请号：US18347297

申请日：2023-07-05

Applicant: ELECTRONICS AND TELECOMMUNICATIONS RESEARCH INSTITUTE

Inventor： Sang-Cheol KIM ,  Hyung-Kook JUN ,  Tae-Ho KIM

IPC: G06N3/045 ,  G06N3/063

CPC classification number: G06N3/045 ,  G06N3/063

Abstract: Disclosed herein are an apparatus and method for distributed processing of a neural network. The apparatus may include a neural network model compiler for segmenting a neural network into a predetermined number of sub-neural networks, two or more neural processing units, and a neural network operating system for abstracting the sub-neural networks into a predetermined number of tasks, performing inference using the multiple neural processing units in a distributed manner in response to a neural network inference request from at least one application, and returning an inference result to the application.

公开(公告)号：US20160085568A1

公开(公告)日：2016-03-24

申请号：US14825409

申请日：2015-08-13

Applicant: ELECTRONICS AND TELECOMMUNICATIONS RESEARCH INSTITUTE

Inventor： Vincent DUPRE ,  Tae-Ho KIM ,  Chae-Deok LIM

IPC: G06F9/455 ,  G06F9/48

CPC classification number: G06F9/4812 ,  G06F9/45558 ,  G06F2009/45579

Abstract: Disclosed is a hybrid virtualization method for an interrupt controller in nested virtualized environment, which can reduce guest interrupt latency. A hybrid virtualization method includes operating a Virtual Machine Manager (VMM) which enables a program executed on a host and a program executed on a guest to be simultaneously used, providing, by the VMM, a virtual interrupt configuration register to the guest, and selectively utilizing full virtualization and partial virtualization. Full virtualization is used when guests which are running simultaneously request the same interrupt. In this case, the interrupt is trapped by the VMM before being injected to the different guests. Partial virtualization allows a guest to directly handle incoming interrupt and avoids expensive traps to the hypervisor to reduce the latency. This virtualization technique can be used by any kind of VMM which is physical hypervisor or virtual hypervisor for processing interrupts of their respective guests.

Abstract translation: 公开了一种嵌套虚拟化环境中的中断控制器的混合虚拟化方法，可以减少客户中断延迟。 混合虚拟化方法包括操作虚拟机管理器（VMM），其使得主机上执行的程序和在客户端上执行的程序被同时使用，由VMM向访客提供虚拟中断配置寄存器，并且选择性地 利用完全虚拟化和部分虚拟化。 当同时运行的客户请求相同的中断时，将使用完全虚拟化。 在这种情况下，中断在注入不同的guest虚拟机之前被VMM所困。 部分虚拟化允许客户端直接处理传入中断，并避免向虚拟机管理程序发送昂贵的陷阱以减少延迟。 这种虚拟化技术可以被任何种类的VMM使用，VMM是物理管理程序或虚拟虚拟机管理程序，用于处理他们各自客人的中断。

公开(公告)号：US20140122937A1

公开(公告)日：2014-05-01

申请号：US13943399

申请日：2013-07-16

Applicant: Electronics and Telecommunications Research Institute

Inventor： Chang-Min SHIN ,  Tae-Ho KIM ,  Chae-Deok LIM

IPC: G06F11/36

CPC classification number: G06F11/3636 ,  H04B7/18506

Abstract: Disclosed herein is a monitoring method and apparatus for an ARINC 653-based operating system. In the monitoring method for an ARINC 653-based operating system according to the present invention, a client system transfers information about a breakpoint to a target system. The breakpoint information is set in the target system. A Micro Controller Unit (MCU) reads a Program Counter (PC). If the breakpoint information matches the PC, an interrupt is caused. A function contained in the breakpoint information matching the PC is executed.

Abstract translation: 本文公开了一种用于基于ARINC 653的操作系统的监视方法和装置。 在根据本发明的用于基于ARINC 653的操作系统的监视方法中，客户机系统将关于断点的信息传送到目标系统。 断点信息在目标系统中设置。 微控制器单元（MCU）读取程序计数器（PC）。 如果断点信息与PC匹配，则会引起中断。 执行与PC匹配的断点信息中包含的功能。